Skip to main content

Full text of "A History of Greek Philosophy Vol. 5"

See other formats


A HISTORY OF 
GREEK PHILOSOPHY 
VOLUME V 
A HISTORY OF 
GREEK PHILOSOPHY 
BY 
W. K. C. GUTHRIE 
VOLUME V 
THE LATER PLATO 
AND THE ACADEMY 
CAMBRIDGE UNIVERSITY PRESS 
CAMBRIDGE 
LONDON · NEW YORK · MELBOURNE 
Published by the Syndics of the Cambridge University Press 
The Pitt Building, Trumpington Street, Cambridge CB2 irp 
Bentley House, 200 Euston Road, London nwi 2DB 
32 East 57th Street, New York, ? ? 10022, USA 
296 Beaconsfield Parade, Middle Park, Melbourne 3206, Australia 
© Cambridge University Press 1978 
First published 1978 
Printed in Great Britain at the 
University Press, Cambridge 
Library of Congress Cataloguing in Publication Data 
Guthrie, William Keith Chambers, 1906- 
A history of Greek philosophy. 
Includes bibliographies 
CONTENTS: v. 1. The earlier Presocratics and Pythagoreans.—v. 2. The Presocratic 
tradition from Parmenides to Democritus.—v. 3. The fifth-century enlightenment [etc.] 
1. Philosophy, Ancient - History. I. Title. B171.G83 182 62-52735 
ISBN ? 521 2C003 2 
Preface page xiii 
List of Abbreviations xv 
I Cratylus 
Date; dramatic date; characters I 
The dialogue 5 
Comment 16 
general; conversation with Hermogenes; a fallacy of 
division? Hermogenes and Protagoras; essence and 
form; the etymologies; the right relation between 
names and reality; what is meant by correctness of 
names 
Additional note: an ideal language? 31 
II Parmenides, Theaetetus, Sophist, Politicus 
Introduction 32 
A) Parmenides 33 
Introduction (theistic interpretations); date; dramatic 
date; setting and characters 
Part One A263-35 d) 36 
Zeno's argument countered by the doctrine of Forms; 
Parmenides s questions and objections; (i) of what 
things are there Forms? (ii) what shares in a Form 
must contain either the whole of it or a part; (iii) 
first regress argument: the largeness of the Large; 
(iv) can the Forms be thoughts? (v) second regress 
argument: Forms as patterns or paradigms; (yi) the 
Forms unknowable to us and we to God 
Conclusion on Part One 50 
Transition to Part Two 52 
Part Two A370-660) 54 
Conclusion 57 
B) Theaetetus 61 
date; dramatic date; characters; prefatory  
conversation; introduction to main dialogue; the question, 
what is knowledge? additional note on  
exemplification and definition; plan of the enquiry; introductory 
conversation 
A) Knowledge as perception A51 d-86e) 73 
what is included in aisthesis? Protagoras and his 
'secret doctrine'; the cleverer theory of sensation; 
status of the sensible world; dreams and  
hallucinations; examination of the theory that knowledge is 
perception: (i) return to Protagoras; (ii) foreign 
languages and unlearned letters; (iii) memory; (iv) 
the 'knowing and not knowing' dilemma; (y) back 
again to Protagoras: the defence; (vi) criticism of 
the defence; (vii) final refutation of Protagoras 
Digression: the philosopher and the practical man 
A72C-77C) 89 
summary; the lesson of the Digression 
Excursus: evil and its sources 92 
(i) evil as a negative conception; (ii) evil due to body 
or soul? (iii) are there Platonic Forms of evils? 
Return to A) 100 
final attack on the theory of total flux; final 
disproof of the identification of sensation with 
knowledge 
B) Knowledge as true judgement (doxa) A87 b- 
201 c) 103 
but is false judgement possible? (j) false judgement 
as mistaking one thing for another; (ii) false  
judgement as thinking what is not; (iii) false judgement as 
'other-judgement'; (iv) false judgement as the 
misfitting of a perception to a memory: the mind as a 
wax tablet; (v) knowledge potential and actual: the 
aviary 
C) Knowledge as true judgement with a logos 
B010-iob) 114 
the theory dreamed by Socrates; three possible 
meanings of logos 
Conclusion 120 
C) Sophist 122 
introductory remarks; definitions: the angler and the 
Sophist; comment on definitions 1-6; diairesis; 
seventh and final definition: the Sophist as illusion- 
maker 
The status of 'what is not' and the criterion of 
being B37a-48e) 135 
(a) the Sophist's reply; (b) from the unreal to the 
real; (c) materialists and idealists: the criterion of 
being; 
motion has a place in the real world; problem of 
motion and reality; interrelationship of Forms; five 
of the greatest Forms; Parmenides refuted; speech 
and thought: the nature of falsehood; return to 
dichotomy: the Sophist finally captured; the Sophist 
and the Forms 
Additional notes 161 
(i) ' the Logos has its birth through the interweaving 
of Forms with each other9 B59d); (ii) Republic 5 
and the Sophist 
D) Politicus 163 
Introduction, outline and general remarks 163 
A) Logic and method 166 
(a) collection and division; (b) the two types of 
measurement; (c) the use of paradigm 
B) Forms in the Politicus 175 
C) The myth 180 
D) Political theory 183 
(a) Politicus and Republic; (b) rule by force or 
consent; (c) the role of law in governments (d) final 
isolation of the statesman; (e) the essence of 
statesmanship 
E) Ethics and psychology 191 
Conclusion 192 
Appendix: Elements of the myth 193 
III Philebus 197 
date and characters; the concepts of pleasure and 
good; subject and scope; the argument; the one-and- 
many problem; dialectical solution; fourfold 
analysis of everything; the cause: cosmological and 
teleological arguments; psychology of pleasure, pain 
and desire; false pleasures; are there any true 
pleasures? pleasure—pain compounds; puzzles; true 
pleasures; pleasure as process and means; analysis 
of knowledge; composition of the mixed life; pleasure 
loses second prize; the five possessions; the  
philosophy of the Philebus/ conclusion 
IV TlMAEUS AND CRITIAS 241 
Introduction 241 
influence; date and characters 
Framework and purpose 244 
Atlantis {Tim. 20d-25d, Crit. 108e-21c) 247 
The ' probable account' 250 
Maker, Model and Material 253 
maker; model; relation of maker to model; material 
What exactly enters and leaves the Receptacle? 269 
What is the cause of pre-cosmic motion? 271 
Necessity 272 
Creation of cosmos B9d-34b) 275 
why it was created; uniqueness of the cosmos; body 
of the cosmos 
Construction of the primary bodies E3 c—57d) 280 
geometrical basis of matter; transformation of the 
primary bodies; fifth figure and fifth body; the 
remoter principles: geometry and physics; particles 
vary in si\e; perpetual motion and warfare of the 
primary bodies; motion demands both mover and 
moved; five worlds? 
Soul of the cosmos C4b-36d) 292 
Time and creation 299 
Creation of living creatures: nature and fate of the 
human soul C9e-42c) 305 
making and destiny of human souls; the infant soul 
and the cause of error 
Additional notes 311 
(/) the status of eros; (ii) extra-terrestrial life? 
Necessity and design in the natures of men F1 c- 
9od) 313 
teleological explanation; physiology based on 
physics; sensible qualities; body and soul; pleasure 
and pain; diseases of body and soul 
Appendix: the narrative order 319 
V Laws 321 
Introduction 321 
authenticity and date; characters and setting; plan 
if chapter 
A) Introductory conversation (bks 1-3) 325 
aims and methods of education, with special 
reference to the use of drink; unity and multiplicity 
of virtue; the lessons of history; need for a mixed 
constitution 
B) The city of the Laws 332 
was it intended to be realised in practice? status and 
function of laws: the lawgiver as educator; the role of 
punishment; theory and reality 
C) Life in Plato's city 341 
population; public before private weal; private 
property; the four classes; trade and labour; 
education; slavery; daily life in Magnesia; contact 
with the rest of the Greek world; women; sexual 
morality and procreation; conclusion: the ideal 
citizen 
D) Religion and theology 357 
state religion and ethics; personal beliefs; theology 
Additional note: Is soul something created? 366 
E) Preservation of the laws: the Nocturnal 
Council 368 
F) The Laws in Plato's philosophy 375 
general; the attitude of the law to the Socratic 
dictum that no one does wrong voluntarily; the Laws 
and the theory of Forms 
G) Conclusion 381 
VI Doubtful and Spurious Dialogues 383 
Introduction; Epinomis; Second Alcibiades; Clito- 
phon; Hipparchus; Minos; Rivals; Theages; 
Axiochus; Eryxias; Demodocus and Sisyphus; On 
Justice and On Virtue 
VII Letters 399 
The philosophical section of the Seventh Letter 402 
VIII Plato's 'Unwritten' Metaphysics 418 
Introduction: the modern thesis 418 
A glance at the evidence 423 
Content of the unwritten doctrine 426 
the archai of the Forms: Indefinite Dyad and One- 
Good; the Forms as numbers; Forms and numbers 
the same, or numbers the archai of Forms? did Plato 
limit the numbers to ten? the generation of numbers; 
general scheme; were Plato s metaphysics monistic 
or dualistic? 
IX Postscript To Plato 443 
X Plato's Associates 446 
Eudoxus 447 
life; mathematics; astronomy; ontology; Eudoxus 
and the Forms; geography and ethnology 
Speusippus 457 
life; ontology; theology; biology; philosophical 
method and epistemology; psychology; ethics: 
Speusippus on pleasure 
Xenocrates 469 
life and character; writings; being and knowledge; 
the chain of being; the Forms; theology: gods and 
daemons; cosmology and physics; indivisible lines, 
atomic bodies, parts and wholes; method and logic; 
psychology; ethics 
Heraclides Ponticus 483 
life; writings; astronomy and cosmology; physics; 
theology; the soul; pleasure; additional note: the 
nature of sound 
Others 490 
Bibliography 493 
Indexes 
I Index of passages quoted or referred to 515 
II General index 528 
III Index of Greek words 539 
The device on the front cover is the head of Plato 
from a herm in the Staatliche Museen, Berlin 
PREFACE 
There is no break in subject-matter between this volume and its 
immediate predecessor, and their division is purely a matter of physical 
convenience. Perhaps therefore they might better have been called 
1 Volume iv parts ? and 2', but since they are in fact two separate and 
fairly bulky volumes that did not sound right either. It does mean 
however that what was said in the previous preface about the general 
approach adopted in this study of Plato, its aims and methods, applies 
equally to this second half and need not be repeated. No interpreter of 
Plato can feel fully satisfied with his work, if only because of the 
inevitable choice, whether to make the main part of the exposition an 
analysis and appreciation of separate dialogues or a synthetic or 
systematic treatment by subjects. I do not regret the decision for which 
I gave reasons in my last preface, but as I also admitted, there are  
drawbacks in either method. In the present volume (ch. vin) I have tried to 
do justice to the modern school of interpreters who see Plato as from 
his early days a systematic thinker with a settled doctrine of first 
principles, orally, even secretly, expounded, which, though we can now 
only glimpse it through the veil of Aristotelian and later criticism and 
comment, must be assumed as the unwritten background to every stage 
of his written work. On these premises it is of course wrong to hold 
back the esoteric teaching until after the dialogues, but as will appear, 
I cannot regard the thesis as established beyond question, whereas on 
the other hand I do perceive, and hope I have brought out, a number of 
lines of genuinely philosophical development in the dialogues  
themselves. (It is this development which I hoped would save volume iv 
from appearing, as it did to one critic, more like a series of monographs 
than parts of a continuous history.) On the question of arrangement see 
also my 'Postscript' (ch. ix). 
I should like to express my thanks to friends who have read some of 
my chapters on the dialogues and made valuable suggestions, many of 
which I have adopted to the great improvement of the chapters  
concerned. Vol. iv ch. vii {Republic) was read by Sir Desmond Lee, and in 
xiii 
Preface 
the present volume ch. ?? (Philebus) by Professor Sandbach and 
Dr G. E. R. Lloyd. Dr Lloyd also read ch. iv (Timaeus) and Dr T. J. 
Saunders ch. ? (Laws). To Dr Saunders in particular I owe a number of 
useful references which had escaped me. For these as for all other  
chapters, however, I remain solely responsible, especially as I did not adopt 
every suggestion offered. To Miss ?. ?. Gorse I am indebted for three 
things: her impeccable typing, her classical education, and a friendship 
extending over many years. 
Unattributed references to 'vol. i' etc. refer as before to the earlier 
volumes of this work. 
I should like to correct a somewhat elusive misprint in the preface to 
the first impression of vol. iv. On p. xv, 1. 10, for 'effect' read 'defect*. 
I also apologize for the blank space on p. 4, n. 1. The reference should 
be to pages 63 f. 
CAMBRIDGE W. K. C. G. 
DECEMBER 1976 
xiv 
LIST OF ABBREVIATIONS 
Most works cited in abbreviated form in the text will be easily  
recognizable under the author's or editor's name in the bibliography. It may 
be however helpful to list the following: 
PERIODICALS 
AGPh 
AJP 
APQ 
BICS 
BJPS 
cj 
CP 
CQ 
CR 
GandR 
GGA 
HSCP 
IPQ 
JHI 
JHP 
JHS 
JP 
PAS 
PCPS 
PR 
Phron. 
PQ 
RCSF 
REG 
???? 
Archiv fiir Geschichte der Philosophie 
American Journal of Philology 
American Philosophical Quarterly 
Bulletin of the Institute of Classical Studies (London) 
British Journal for the Philosophy of Science 
Classical Journal 
Classical Philology 
Classical Quarterly 
Classical Review 
Greece and Rome 
Gbttingische Gelehrte Antigen 
Harvard Studies in Classical Philology 
International Philosophical Quarterly 
Journal of the History of Ideas 
Journal of the History of Philosophy 
Journal of Hellenic Studies 
Journal of Philosophy 
Proceedings of the Aristotelian Society 
Proceedings of the Cambridge Philological Society 
Philosophical Review 
Phronesis 
Philosophical Quarterly 
Rivista Critica di Storia della Filosofia 
Revue des Iitudes Grecques 
Transactions of the American Philological Association 
XV 
Abbrevations 
OTHER WORKS 
(Full particulars are in the bibliography) 
CGF Comicorum Graecorum Fragmenta, ed. Meineke 
DK Diels-Kranz, Die Fragmente der Vorsokratiker 
D.L. Diogenes Laertius 
KR G. S. Kirk and J. E. Raven, The Presocratic Philosophers 
LSJ Liddell-Scott-Jones, A Greek-English Lexicon, 9th ed. 
OCD Oxford Classical Dictionary 
OP Oxyrhynchus Papyri 
Posidonius, 
EK The fragments of Posidonius, ed. Edelstein and Kidd. 
PS G. Vlastos, Platonic Studies 
RE Realencyclopadie der classischen Altertumswissenschaft, ed. 
Wissowa, Kroll et al. 
SPM Studies in Plato's Metaphysics, ed. R. E. Allen 
SVF Stoicorum Veterum Fragmenta, ed. von Arnim 
TGF Tragicorum Graecorum Fragmenta, ed. Nauck 
Note: The dialogues known in England as Republic and Politicus are in 
some countries called Politeia and Statesman (in the language of the 
country) respectively. Non-English readers should be warned that 
the abbreviation Pol. indicates the latter work. 
xvi 
I 
CRATYLUS1 
If you are on your guard against taking names too seriously, you will 
be richer in wisdom as you grow old. Plato, Pol. 261 e 
Date. The placing of this dialogue immediately after the Republic is not 
intended as a pronouncement on its date, which, like its purpose, has 
been a matter of lively debate. Earlier critics (e.g. all five in Ross's 
table, PTI3) thought it an early dialogue, before Phaedo, Symposium, 
Phaedrus and Republic, and von Arnim's stylistic studies made him date 
it around 390, before Plato's first Sicilian visit, though otheis (see Ross, 
ib. 4-5) had seen affinities with later dialogues. Ross himself argued in 
1955 for an early date, and Taylor thought it earlier than any of the 
'great dramatic group', even the Protagoras. But since the fifties the 
argument from apparent affinities in content with the so-called ' critical 
group' {Parmenides, Theaetetus, Sophist, Statesman) has won much 
more favour, though still without unanimity. Runciman A962) places 
it 'with reasonable confidence' before that group on grounds both of 
style and less sophisticated treatment and thinks Phaedrus is later, Luce 
A965) takes it as preceding Phaedo and Republic and Brentlinger A972) 
still puts it before Symposium, Phaedo and Republic, as in 1931 did 
Meridier. On the other hand Owen A953) thinks the argument at 
439d8-9 'alone would vindicate its place in the critical group', Kirk 
A951) and Allan A954) put it contemporary with Theaetetus, and 
Schadewaldt A971) also argues for a fairly late date, as an immediate 
preliminary to the critical group. In 1953 Jowett's editors disputed his 
comparatively early dating of the Cratylus and emphasized its affinities 
with the later dialogues.2 
1 A descriptive bibliography of works on the Cratylus 1804-1972 will be found in Derbolav, 
Sprachphil. 1972, 234-308. 
a RefT. not supplied in the text are: Ross, R. Int. de Phil. 1955; Taylor, ????^; Runciman, 
PLE 2 and 129; Luce, Phron. 1965, 21 and 36; Meridier in his Bude ed., 46; Owen, SPM 323 
n. 3; Kirk, AJP1951, 226; Allan, AJP 1954, 272; Schadewaldt, Essays Merlan 3-11; Brentlinger, 
AGPh 1972, 116 n. 1; Jowett, Dial.4 in, 10 n. 1. For a conspectus of views before 1941 see 
Leisegang, RE 40. Halbb. 2428. He himself, like Meridier and Wilamowitz, found it impossible to 
2 
I 
GHG 
Cratylus 
Lest it appear that the arguments for an early or late date depend on 
giving chief weight to style or content respectively, it should be added 
that an important argument for the earlier date concerns the stage which 
has been reached in the doctrine of Forms. Thus Meridier, Ross (PTI 
18-20) and Luce (Phron. 1965, 36) have maintained that they are not 
yet fully transcendent or' separated' (Aristotle's word) from particulars, 
a view which would of course put the Cratylus before the Phaedo. 
The above selection of opinions will suffice to justify Crombie's 
assessment of the Cratylus as 'a dialogue whose date must be left  
uncertain' (EPD 11, 323). More even than most, it is a unique and self- 
contained whole. 
Dramatic date. It is usually thought (see e.g. Meridier 46) that the 
dialogue contains no indication of when the conversation was supposed 
to have taken place, but Allan has argued {AJ ? 1954,272-4) that it was 
during the last year of Socrates's life. 
Characters. Apart from this dialogue, we know of Cratylus only from 
Aristotle's statements that Plato was acquainted with him in his youth, 
and learned from him the doctrine that everything was in flux, which at 
some time he held in a more extreme form than that taught by Heraclitus 
himself. (See vol. 111, 201.) Plato too speaks of him as a Heraclitean 
D37a 1, 440d-e) and even attacks Heracliteanism in the extreme form 
in which Cratylus himself (according to Aristotle) held it: if all things 
are in flux, they cannot even be spoken of D39 d). Scholars have found 
difficulty in reconciling this with all the beliefs ascribed to him in the 
dialogue, and some have sought to avoid it by assuming that Plato is 
only using his name to make a veiled attack on someone else. Antisthenes 
was a favourite guess in the past, but is less popular now.1 Though in 
separate Crat. from Euthyd. Nakhnikian has argued persuasively for the priority of Crat. to Tht. 
from their respective treatments of Protagorean and Heraclitean views (?. of Metaph. 1955-6, 
308 f.). Latest of all, Kahn (Exegesis 154) in 1973 agrees with Ross in placing it near the beginning 
of the middle group. 
1 'In fact the Antisthenes-theory is almost dead' (Kirk, AJP 1951, 238). A useful list of reff. is 
given by Levinson in R. of Metaph. 1957-8, and a summary of those in favour by Me>idier D4k), 
whose sensible conclusions should be noted. Since Levinson mentions Derbolav as supporting it, 
it is fair to say that in his later book (SprachphiL 1972, 30f.) he concludes that all attempts at 
identification rest on such scanty evidence that to decide between them is to act on faith rather 
than knowledge. 
2 
Cratylus 
general highly suspicious of such conjectural identifications,1 I have 
tried to show in vol. ?? (p. 215) that the central theory of Plato's 
Cratylus, that names have a natural affinity with their objects, was also 
upheld by Antisthenes, as, certainly, was the impossibility of false 
speaking. The importance which he attached to language is indicated 
by his pronouncement that the basis of education was the study of 
names.2 Since the nature and use of words was a favourite topic of 
discussion among the Sophists (vol. 111, 205 f.), there were probably 
more than one champion of each of the opposing views. Another  
suggestion is that the etymological theories of 'Cratylus' are directed 
against Plato's own gifted pupil Heraclides Ponticus. This was put 
forward by Warburg in 1929, but as Meridier says, it 'repose sur une 
base des plus fragiles'.s Protagoras has also had his turn,4 and is  
mentioned in the dialogue itself as an expert on 'the correctness of names', 
whose central doctrine identifying appearance with reality is rejected by 
Hermogenes C91c, 385e-86c). It was a leading theme of vol. 111 that 
the Sophists shared a common scepticism resting on a plausible  
interpretation of Heraclitus's flux-doctrine.5 At the same time they were 
fascinated by the compulsion of Eleatic logic, as is plain from Gorgias's 
use of purely Eleatic arguments to maintain the equally paradoxical 
thesis that nothing exists (vol. in, i93ff.)> and their thesis of the 
impossibility of falsehood seems to have rested both on the Heraclitean 
assertion of the identity of opposites (vol. 111, 166,182 n. 2) and on the 
Parmenidean dictum that 'what is not' cannot be uttered. For their 
purposes Heraclitean and Eleatic doctrine were at one in 'abolishing the 
criterion' for any comparative assessment of judgements about the 
1 See vol. in, xiv, 3iof., 323, 347. 
1 Vol. ??, 209-11, cf. Crat. 383 et al. (natural Tightness of names), 429 d (impossibility of 
falsehood). 
3 Bude* ed. 41, where reff. for the thesis and its critics will be found. It is an odd coincidence 
that the father of Heraclides should have been called Euthyphro (Heraclides fr. 3 Wehrli, where 
see W.'s note). See also Skemp, TMPLD, i{. 
4 First argued by Stallbaum. See Derbolav, Sprachphil. 30 and 297. 
5 Though H. himself would not have drawn the same epistemological conclusions, for 
????? ??? was not the whole of his teaching. Cf. frr. 1 and 2 (the common logos and the folly of 
acting 'as if each had his own private wisdom'), fr. 107 (the senses bad witnesses if not checked 
by the psyche), fr. 114 (the need for voos; the one divine ????* which feeds human ?????). See 
for these vol. I, 425, 415, and cf. vol. ill, 185. Jackson (Praelections 17-19) has some judicious 
remarks on the question whether the theory of the natural Tightness of names goes back to 
Heraclitus himself. 
3 2-2 
Cratylus 
sensible world and human affairs.1 It is a reasonable conclusion that 
Plato found Cratylus the Heraclitean a suitable character through which 
to criticize prevailing beliefs of the Sophists about the relationship 
between words and reality. How far the historical Cratylus shared their 
linguistic doctrines we cannot be sure, but at least it is unjustified to say 
with Warburg and Heinimann that because the Sophists were not pure 
Heracliteans, Heracliteanism and etymology were unconnected until 
Plato himself combined them in the person of Cratylus. 
ADDITIONAL NOTE: CRATYLUS, HERACLITUS AND THE 
CORRECTNESS OF NAMES 
For the view just mentioned see Heinimann, N. u. PL 54. In the exchange of 
views between Kirk and Allan in AJP 1951 and 1954, I do not find either 
entirely convincing. It is difficult to believe with K. (p. 244; cf. Allan 281 f.) 
that Aristotle's accounts of C.'s Heracliteanism (MetapL 989329 and 
1078 b 9) are taken from Plato (though it was suggested in 1829; see Derbo- 
lav, Sprachphil. 283), and A.'s hypothesis of two stages in his development is 
precarious. K.'s view is influenced by his belief that Plato regularly  
misrepresented Heraclitus, on which see vol. 1, 488-92. His argument that in the 
dialogue C. only welcomes Heracliteanism because it supports his belief in 
the natural correctness of names, not vice versa, is weakened by the fact that 
C. was a historical person known to be a Heraclitean. In his book on  
Heraclitus (HCF n8f.) K. actually argues that H. himself did believe that names 
give some indication of the nature of their objects and bear an essential 
relation to it. The crucial passage is fr. 48 (????-????). (Others quoted by 
Heinimann and Kirk offer less compelling evidence.) Contrary to what 
Heinimann says (ox. 55), this does not deny the view of names attributed to 
C, which was not that names commonly in use are correct, but that they are 
attached to things by convention C83 a), being either the name of something 
else or mere noises. Heraclitus with his example of the bow may have meant 
the same, but more probably he used it to illustrate his doctrine of the 
identity of opposites: life and death are the same (fr. 88; vol. 1, 445 f.). 
HermogeneSy son of Hipponicus and brother of Callias the wealthy 
patron of Sophists (vol. 111, 31 and iv, 216), was a close associate of 
Socrates who according to Plato was one of those present during his 
last hours in prison. Xenophon says he had also been at the trial, and 
had previously tried to persuade Socrates to give some thought to his 
1 On 'abolition of the criterion' as a mark of the Sophists, see Gorg. fr. 3 DK (vol. in, 195 f.). 
4 
The dialogue 
defence. He appears again as a participant in Xenophon's Symposium. 
Diogenes Laertius called him a follower of Parmenides, but this is 
doubtful, and perhaps an inference from his appearance here as the 
opponent of Cratylus.1 Otherwise nothing is known of his views apart 
from what is here attributed to him by Plato, who represents him as a 
young man with no great aptitude for philosophical discussion. Both 
the respondents are depicted as younger than Socrates D29b, 440d), 
but Cratylus shows much more self-assurance and tenacity in  
maintaining his opinions. 
The dialogue1 
{Direct dramatic form) 
Hermogenes and Cratylus have been arguing about the status of names3 
in terms of the current nomos-physis antithesis,4 and agree to refer the 
dispute to Socrates. Hermogenes holds that they are merely  
conventional labels imposed by agreement or custom (nomos) and changeable 
at will, whereas Cratylus, he says, claims that everything has a naturally 
correct name, the same for Greeks and foreigners, irrespective of those 
in current use. Hermogenes cannot understand what he means, and he 
refuses to explain himself. Socrates as usual disclaims knowledge but is 
willing to go into the matter with them and starts his questions. The 
rest of the dialogue falls into two parts, carried on with Hermogenes 
and Cratylus respectively. 
1 Phaedo 59b, Xen. Apol. 2 and 3, Mem. 4.8.4, D.L. 3.6. That the last is an inference from 
Crat. was suggested by Natorp, RE vin, 865. 
1 A brief indication of the contents has been given in vol. 111, 206-10. 
3 I shall in general keep 'name' for the Greek ?????, even though 'word' would sometimes 
be a more natural translation. As Robinson said (Essays 100), there is no exact English equivalent. 
Examples in Crat. include proper names, nouns, adjectives and even adverbs D27 c), and in 
Soph. B62a) (though not always) they are distinguished from verbs. Thus P.'s use approximates 
more to Mill's than to present usage. (Mill, Logic bk 1 ch. 2.) M. Roth has a good discussion, in 
an unpublished Illinois dissertation on the Crat. A969), 33-6. 
4 In vol. in, 206 n. 2,1 followed Fehling in saying that the contrast is not between ????? and 
?????. (So also Robinson, Essays noff.) This is literally true, but not really important. The 
contrast is between names which are formed ????? (i.e. although man-made, they possess a 
natural Tightness to which the makers conformed) and ????? (only). I think therefore that  
anything known about ????? versus ????* on this matter is relevant, e.g. Simpl.'s statement (in Catt.y 
Schol. Bekk. 43 b 31) that the Pythagoreans said names were ????? not ?????. 
5 
Cratylus 
(?) Discussion with Hermogenes: there is a natural correctness of names 
C85a-427d). Hermogenes repeats his belief that what even a single 
individual chooses to call something is as much its name as any other. 
If I call' man' what everyone else calls' horse', no one can challenge me, 
though its 'public' name is different. But he admits that speech may be 
true or false, and it follows, says Socrates, that the components of 
speech, including names, must be true or false.1 If each object has as 
many names as anyone chooses to give to it, so long as he gives them, 
Protagoras must have been right in saying that things themselves are 
only what they appear to each one of us. They agree in rejecting  
Protagoras, but the only alternative is that every existing thing has a stable 
nature or essence (?????) of its own, irrespective of our beliefs about it. 
Actions too (Socrates proceeds) are realities, whose character is 
shown by the fact that we can only perform them as, and with the 
instruments which, their nature demands, not according to our own 
whim. You cannot cut with a box of matches or light a fire with a knife. 
Speech is an action, performed with words which are its proper  
instruments just as shuttles are for weaving. They have a dual function, 
(a) communication (lit.' teaching one another'), and (b) the  
distinguishing or differentiation of one real thing from another.2 Any tool, to 
perform its function properly, must be made by a skilled worker, and 
names are no exception. Since (as Hermogenes has said) they are the 
product of nomos, their maker must have been a lawgiver (nomothete), 
the rarest of all human craftsmen.3 
Taking the analogy a step further, when a carpenter makes a shuttle, 
he has in view its function, and if one breaks under his hand he does not 
take the broken one as his model but' the sort of thing which is fitted 
by nature to be a shuttle', which we may call the ideal shuttle ('what a 
shuttle is in itself, 389 b). Within their generic function shuttles serve 
1 The truth or correctness of a name, as S. says later, depends on its revealing the nature of its 
object (???? ???? ?? ??????, 428 e). 
% S. is continuing the analogy from weaving: as the shuttle separates (?????????) the threads, 
so names separate the realities which they name. 
3 There have of course been plenty of lawgivers, but S. is speaking of the expert, whose names 
will correctly distinguish the essence of their objects, and as we know, for P. this is none other 
than the philosopher. S. is running rings round poor H., who will not of course see this point, and 
apart from that, he takes nomos to mean law, when H. obviously used it in its other sense of 
custom. Goldschmidt's defence (Essai6i(.)y that the two concepts were indissoluble in the Greek 
mind, fails. Nomos as custom is not the work of a nomothete. 
6 
The dialogue 
specific ends, for fine or coarse weaving, in wool, linen or other 
materials. Each must be given both the generic form and the character 
{physis) suited to its specific purpose. Similarly the master name-giver 
must put the name naturally formed for each purpose into sounds and 
syllables as well as keeping an eye on the ideal name. Different syllables 
may be used, as the same tools (say two hammers) are made out of 
different pieces of metal. Provided they are correct in form, they serve 
their purpose equally well, and so do names whether in Greek or 
another language. 
The man who will know the proper form for an instrument, under 
whose instructions the maker must work, is the user—weaver, musician, 
or in the case of a rudder, a sailor; and the user from whom the name- 
maker must take his instructions is the dialectician, the skilled asker and 
answerer of questions. Naming is no light undertaking. Things, as 
Cratylus says, have names by nature, and their giver must look to the 
natural name of each thing and be able to put its form into letters and 
syllables. 
Hermogenes would be happier about this if he knew what 'natural 
correctness of names' meant. He has forgotten that Socrates is just a 
fellow-enquirer, and 'correctness of names' is the province of Sophists. 
However, even Homer and the other poets can teach us something. 
Homer speaks of different names given by gods and men,1 and  
presumably the gods know the right one. Without meddling in such high 
matters, even we can judge between the two names he gave Hector's 
son, Scamandrios and Astyanax.2 Both are Greek, and since Hector 
means 'holder (or sustainer)' of the city and Astyanax 'lord of the 
city', obviously Astyanax is right, on the principle that son resembles 
father, as a lion's cub is called a lion. This does not always work. Even 
in nature there are freak births, and a pious man may have an impious 
son. In that case Theophilos ('God-friend') or Mnesitheos ('mindful- 
of-God') are incorrect names for him. The correct one would signify 
1 As examples S. quotes //. 20.74, 24.291 and 2.813 f. 
1 II. 6.402 f. S. says that because the Trojan men called him Astyanax, it must have been the 
women (the sillier sex) who called him Scamandrios! Unless there was an alternative version now 
lost, he was relying on H.'s imperfect memory of Homer to pull his leg, for in Homer Hector 
himself called the boy Scamandrios. The derivations of Hector and Astyanax are of course 
correct, and in fact most Greek proper names have a transparent meaning. 
7 
Cratylus 
the opposite. Others are merely examples of wishful thinking, like 
Eutychides ('son of good fortune'). (This example comes to life when 
we read a funerary inscription in which a dead man called Eutychides 
complains that he was wrongly named. See Luce, CQ 1969, 225.) On an 
earlier point, we see how little the material constituents matter, for 
Hector and Astyanax have scarcely a letter in common. Similarly 
Iatrokles ('famous physician') and Akesimbrotos ('healer of men'), 
though so different in sound, are both correct names for doctors.1 
Socrates now proceeds to show, by a torrent of etymologies,2 how 
other names or words also reveal the nature of their objects. He, the 
ignorant, has been filled with a miraculous wisdom, doubtless caught 
from the inspired seer Euthyphro, to whom he has earlier been listen- 
ing.3 Tomorrow he will exorcize the spirit 'through some priest or 
Sophist', but today he will exploit it. Beginning with some basic words 
—god, man, soul,4 body—he passes to the gods, of whom he avers that 
a claim to know their true names would be blasphemous: we can only 
say what was in the minds of men when they gave them their names. 
In spite of this, he proceeds as if the known names do reveal the gods' 
real natures, e.g. Demeter means 'the giving mother', Pluto means 
'wealthy' and his other name Hades 'knowing'. There follows a purely 
Platonic digression on Pluto's philosophic nature: he consorts only with 
souls that are freed from the body and its insane desires, and keeps them 
spellbound by the riches of his wisdom. Hephaestus is 'obviously' 
Phaestus ('lord of light'), the first syllable being a mere extraneous 
addition. 'Probably—until you get another idea', says Hermogenes. 
' Well, to prevent that, ask me about Ares.' 
1 For different languages (cf. 389d-c>oa) S. gives no examples, but 'Zimmermann' and 
4 Carpenter* might be said to reveal the identical nature of their nominates as workers in wood (the 
latter originally as a maker of chariots). As surnames, S. would say, they are only correct for 
families pursuing this trade. 
1 These obviously cannot be fully recounted in an English summary. A few will be included as 
illustrations, and for the whole collection I would refer (though without necessarily agreeing with 
his conclusions) to Boyance in REG 1941, who goes through them in detail to prove his point 
that they are to be taken seriously, especially for their religious significance and as evidence of 
P.'s debt to the Pythagoreans. See also Dummler, Akad. 131??., Haag, P.'s K. ch. 4, Goldschmidt, 
Essai 185-99. 
3 P.'s real opinion of Euthyphro has appeared in the Euthyphro. See especially vol. iv, 107 f. 
4 After very reasonably connecting ???? with ???????, he abandons this for a wildly  
implausible etymology on the grounds that Euthyphro would be more likely to approve it. 
8 
The dialogue 
But the gods are a dangerous subject. Let Hermogenes 'learn the 
mettle of Euthyphro's horses'1 in another sphere. So they turn to the 
heavenly bodies, seasons and elements.2 On 'fire' (m/p) the muse of 
Euthyphro deserts Socrates, and he falls back on the idea that it might 
be a loan-word from a non-Greek source, possibly Phrygian.3 (This is 
invoked again for the intractable kakon ('bad') at 416a and dphelimon 
('beneficial') at 421c, but dismissed at 425e-26e, along with the 
hypothesis of corruption through age, as merely an ingenious device to 
escape the burden of explanation.) 
After this section Hermogenes remarks that Socrates is making great 
progress, and he replies with satisfaction, 'Yes, I do seem to be far 
advanced in this skill, and you will soon have even better reason to say 
so.' They pass to the virtues which, like some of the gods, seem to have 
been named by Heracliteans, for they all have to do with movement and 
flow, e.g. dikaion ('just') is really diiaion ('penetrating') with a k for 
euphony. In this part, Hermogenes says, Socrates seems to be only 
repeating things he has heard, and Socrates replies that he will now try 
to fool him into believing that he is being original. His derivation of 
techne, which involves removing the t and inserting an ? between ch 
and ? and ? and e> seems even to Hermogenes a bit far-fetched. Ah, 
but he doesn't understand. The original names have suffered not only 
from lapse of time but from unscrupulous titivating. People with no 
respect for truth have distorted them for euphony in all sorts of ways 
until no one knows what on earth they mean. If people can add or 
remove letters at will, any name can be fitted to any object, so  
Hermogenes as a wise overseer must check them in the cause of limit and 
probability.4 
1 An adaptation of //. 5.221 f. Perhaps here the steeds which bore E. up to his heavenly 
visions, like Parmenides (vol. 11, 7). 
1 A point of interest for P.'s cosmology was noted by Boyanc? (REG 1941, 147). At 4iob-c 
S. mentions five elements, distinguishing air from aither as in Epin. (981 c), whereas in Tim. E8d) 
aither is a form of air. No one has thought of using this to help in dating the Crat. Neither shall I. 
The distinction is in any case plain enough in Phaedo A09b). See vol. 1, 270-2, and pp. 284f. 
below. 
3 S. says the Phrygian word for fire is very similar to ???, and Meridier noted that in Armenian 
t is 'hur', and the Armenians were thought to be Phrygian colonists. Phrygian did contain words 
milar to, or identical with, Greek, including ????? itself. See Mon. As. Min. Ant. 4, nos. 16, 17, 
16, 239-43. 
4 This invitation to act in direct opposition to his own theory H. accepts with a meek ' I should 
like to' D14?). 
9 
Cratylus 
But he mustn't be too pedantic or he will unnerve Socrates just as he 
is reaching his climax. More astonishing etymologies follow until 
Hermogenes is moved to remark that names become pretty complicated 
under Socrates's hands. That, says Socrates, is the fault of those who 
made them, and he proceeds unabashed. Derivations rain ' thicker and 
faster' as he nears the end, most of them offering remarkable  
confirmation of the Heraclitean view of the world as all flow and movement, for 
they contain these ideas in their roots:1 in fact, he says, the ancient 
namegivers must have been like some contemporary philosophers who 
in their search for reality make so many turns and twists that they get 
giddy, and project the whirl and motion of their own minds on to the 
external world.2 Finally Hermogenes asks about the' really big and fine' 
names like 'truth' and 'falsehood', 'being' and 'name' itself. ('Being' 
—on—has simply lost an u It should be ion, 'going', and Heraclitus is 
right again!) Socrates has 'knocked them to pieces manfully', says his 
admiring partner, but what of short, simple words like ion itself? Well, 
one could always claim foreign origin or irremediable distortion, but 
such excuses are cowardly. A new procedure is needed. 
The problem is this. So far we have explained names by analysing 
them into their elements, but some names are simple and elemental 
themselves. How can we test the correctness of these? The secondary 
(compound) names revealed the nature of their nominates by means of 
the primary. How do the primary names do it? 
If we had no voices, we should, like the dumb, try to indicate things 
by gestures, e.g. lightness or upward direction by raising the hand, 
heaviness or downward by lowering it, miming their nature, and 
similarly with a galloping horse and other animals. Perhaps then a 
name is a vocal imitation of something. This is not to say that to utter 
' Baa' or ' Moo' is to name a sheep or a cow. So far as a thing has sound, 
shape or colour, its imitation is the province of music or painting. But 
besides sensible qualities, everything (including colour and sound 
themselves) has an essence. If one could imitate that through letters and 
1 For ??????????, * to be profitable', S. rejects the vulgar, commercial (and incidentally correct) 
origin in favour of a wonderful theory that it conforms to the Heraclitean canons by meaning 
* swiftest in motion* Di7b-c). 
2 Goldschmidt aptly compares Phaedo 79 c: the soul grows dizzy and confused when it relies 
on the bodily senses, which can only show it what is constantly changing. 
10 
The dialogue 
syllables, then names could make plain what each thing is. The next 
question therefore is whether this is possible. 
The method we must suppose to have been followed by the ancient 
name-givers is this. First the simplest units, letters, are classified into 
vowels, consonants and semi-vowels, and the vowels into their  
subdivisions.1 Then the objects to be named are similarly analysed to see if 
they too can be reduced to elements2 which will show what they are and 
whether they can be referred to types like the letters. The next step is 
to apply the letters to the objects3 according to their resemblance, 
either one to one or combining the letters in syllables. From syllables 
are built nouns and verbs,4 and from them a great and splendid whole, 
the Logos, 'formed by the art of naming or rhetoric or whatever it may 
be, as a living figure is composed by the art of pain ting'.5 
The present task is to split up language once more into its  
components to see whether primary as well as secondary names were 
rightly given. To complete it is beyond us, and the very idea that letters 
and syllables can reveal things through imitation will sound ridiculous, 
but unless it is true the whole theory of' correctness of names' must be 
abandoned. To Socrates his own conjectures sound arrogant and absurd, 
but he gives them faute de mienx. 
In principle, letters imitate basic notions by the movement or shape 
of mouth and tongue in pronouncing them. Thus the rapid vibration 
of the tongue in r suggests motion (so flow in ' river', also ' run',' rush', 
'tremble', and violent actions like 'strike', 'crush', 'break').6 From 
1 Semi-vowels (neither ???????? nor ???????) include liquids and nasals (Haag, P*s K. 12). 
Aristotle's examples are s and r {Poet, I456b26). The subdivisions of vowels are presumably the 
Greek equivalents of a e i ? and ?, and are only mentioned exempli gratia, for there is no reason 
why the consonants should not be similarly specified. (At Phil, i8b-d, where the division of 
letters into vowels etc. is attributed to the Egyptian god Theuth, he is said to have subdivided all 
classes, net vowels only.) 
3 As words to syllables and they to letters. P. uses ???????? both for elements D22a) and 
letters of the alphabet. 
3 424d5~6. Cf. e4~5 ?? ???????? ??! ?? ???????? ?????????. 
4 ???? in this context is a verb, as at Soph. 262d, though at 399a-b ??? ????? is a ????. 
(Cf. 421 ci.) See vol. in, 220f., for this and the exaltation of the Logos (speech). Perhaps one 
should also remember the supremacy of the Logos in Heraclitus's philosophy. 
5 The above is a fairly close paraphrase of the important and difficult passage 424^253. 
(Those of Haag (P,*s K, 12) and Crombie (EPD 11, 376) may be compared.) It may well repay 
further study. 
6 I have suggested English examples for amusement: sometimes a translation from the Greek 
will serve, sometimes not. (Not even S. can bring ??????? itself into line, 426 c.) 
II 
Cratylus 
'breathy' letters (f, j, {) come 'windy' words—'zephyr', 'puff', 
'sizzle' and so on. D and t compress and hold up the tongue, hence 
words like 'stop', 'bind', 'stand'.1 With / the tongue s/ides or s/ips, 
whence these and words of similar import ('level', 'sleek'), whereas # 
arrests its motion, so that the combination indicates stickiness ('glue'). 
A and e are 'big' letters, hence 'large', 'length'; and of course ? is for 
roundness as in 'orb', 'ovum' (Greek ???, egg). 
But what has Cratylus to say to all this? 
B) Discussion with Cratylus: Truth is not to be got from names {428c- 
4oe). Cratylus is delighted with Socrates's ' oracles', by whomsoever 
inspired. Socrates however is suspicious of such a sudden access of 
wisdom, and to test it will go over the ground again. The premise 
stands that correctness of names lies in their power to reveal the nature 
of their objects. Their purpose then is instruction, which is an art, 
practised, as has been said, by lawgivers; and will not some of these be 
better at their profession than others? Cratylus will not admit this: no 
law is better than another,2 and so it is with names. They are either 
correct or not names at all. So if someone calls him Hermogenes he is 
not even uttering a falsehood? No, for false speech is impossible. To 
speak falsely would be to say what is not, and to say what is not is not 
to say anything,3 but only utter meaningless sounds. 
Socrates starts again. Names, we agree, are imitations of their objects. 
Pictures too are imitations, and (a) they may be wrongly attributed: 
one may mistake a portrait of a man for one of a woman. Cratylus 
retorts that the cases are not parallel, but Socrates presses the agreed 
point that both pictures and names are representations. One can say 
'man', pointing to a picture of a man or a woman, and this is what he 
1 S. made the point much earlier C93 d-e) that provided the operative letter is there, the name 
is correct even if others are added. Thus 'beta' is allowable as the name of the letter b. 
2 This sounds an astonishing statement, but (a) Grote (PL 11, 534 n. q) drew attention to 
Minos 317c, where the author makes S. himself claim that a bad law is no real law, but only seems 
so to the ignorant, and to Xen. Mem. 1.2.42-6. (b) It accords with Protagoras's opinion that  
whatever a city thinks right is right for it so long as it thinks so (Tht. 167c, vol. in, 172). (c) Possibly 
C. is speaking as a Heraclitean. Cf. fr. 114: all human laws are nourished by the one divine law. 
3 The argument attributed to Protagoras and Antisthenes (vol. in, 182 n. 2 and 209-11). Its 
ultimate dependence on Parmenides has been adduced to show that our C. was not the Heraclitean, 
but Protagoras and Antisthenes were hardly followers of Parmenides. (Cf. pp. 3f. above.) The 
problem of'saying what is not' is deliberately shelved D29d), and only solved in Soph. 
12 
The dialogue 
means by false attribution, {b) By means of drawing and colour, a 
painter may produce a good or bad likeness. So too one who imitates 
the essence of things through letters and syllables may not get them all 
right, making a bad (inaccurate) name which yet is a name. Cratylus 
sticks to his guns. A name cannot be wrongly written. Either all the 
letters are right or it is not a name at all, for the alteration has made it 
something else. 
That is true in some cases, replies Socrates—a number for instance. 
If one adds to or subtracts anything from ten, it becomes an entirely 
different number.1 An image, on the contrary, must differ in some 
respect from its original. If some god could reproduce Cratylus in every 
detail—his flesh, life and mind—there would be, not Cratylus and an 
image or copy of him [such as a painter or sculptor might make], but 
two Cratyluses. 
Names, then, as copies, cannot be perfect, or they would be  
indistinguishable from their objects, which is absurd. Inappropriate 
letters may be inserted in a name, and names in a sentence, and  
inappropriate sentences may be included in a composition. Yet the subject is 
still being named or spoken about so long as its general stamp (?????) 
is retained, as Hermogenes and Socrates were saying about letters.2 
Cratylus agrees that this is reasonable, but without wanting to quarrel 
about it, still stubbornly denies that a faulty name is a name at all. 
Patiently Socrates starts again from the first premises. A name  
indicates its object, there are compound and simple names, and the latter 
indicate by resembling the object. The only alternative is the view of 
Hermogenes that the name-makers had prior knowledge of the objects3 
and assigned them names by an arbitrary convention which alone 
authenticates them. It would not matter if they had named 'small' what 
is called 'large'. Cratylus is emphatic that the resemblance-theory is the 
right one. 
1 This might be thought perverse, the point being not that instead of ten one might write nine 
or eleven, but that one might distort the name ' ten' (assuming it is correct) by saying or writing 
e.g. 'teen' or 'tine'. But although the word deka for ten occurs in full in the text, we have to 
remember that the Greeks represented numbers by single letters. Thus ? corresponds to our 10. 
Add one letter, a, and it becomes 11. 
2 P. i2 ?. ? above. 
3 Note how S. slips in something that H. never said. At 438a he will make C. say it and thereby 
get him into a corner. 
13 
Cratylus 
Then the letters (elements) of names must also bear this resemblance, 
and we agreed that r represents motion and hardness, / smoothness and 
softness. Now take the Greek word for hardness (sklerotes). In Eretrian 
dialect it ends in r, but we understand each other though s and r have 
different import. Also we understand it as 'hard' although it contains 
an /. ' Well, as you and Hermogenes said, letters get wrongly inserted 
in course of time, but we understand the intent through custom.' And 
what is custom but convention? At the least it means that letters can 
indicate to us an object to which they have no resemblance. No doubt in 
an ideal language names would always resemble their objects, but as 
things are, convention plays a part too. 
Cratylus still insists that the resemblance between names and their 
objects is so close that names are the only source of information and 'he 
who knows the names knows the objects too'. This applies to the 
discovery of new knowledge as well as the communication of acquired. 
But surely, in original research, to take names as a guide to realities is 
dangerous. What if whoever bestowed them did so under an erroneous 
impression of what they stood for? But this on the Cratylean theory is 
impossible: he must have known the truth, or they would not be names 
at all. Besides, look at their consistency: Socrates himself has argued 
that they all express the same world-view. Against this, (a) consistency 
is no guarantee of Tightness if the initial hypothesis is faulty; (b) it is 
doubtful whether the implied outlook is consistent. Words so far 
examined supported the principle of universal flow and motion, but 
others suggest the opposite.1 
Besides, if names are the only source of knowledge, how could the 
first namer make his names with knowledge of the objects? Cratylus 
can only suppose that they were given by some infallible 'power 
greater than human', and therefore must be right. Any others (like those 
suggesting a static world) are not names at all. All very well, but if two 
sets conflict, what are we to do? Names can no longer help us, and we 
need other criteria by which to judge the truth about existing things 
and see which set of names is genuine. It must therefore be possible to 
1 Here S. takes a number of words, e.g. those for knowledge, enquiry, faith, memory, and 
conversely ignorance, licentiousness etc., and thinks up fanciful etymologies to show that the 
' good' words are derived from ' standstill',' rest',' stopping the flow' and so on, and the bad from 
'going with the god', or 'following realities'. Clearly their inventor was no Heraclitean! 
14 
The dialogue 
learn of realities otherwise than through names, and if possible, surely 
also best. That is, we should understand them directly, through  
themselves, or each other where they have affinities. Things unrelated to 
them cannot signify them.1 Names at the best (as all have agreed) are 
copies of realities, and it is more enlightening to learn from an original 
both about itself and the accuracy of the copy than to learn from a copy 
its own success as a likeness as well as the original it represents. How 
this is to be done—how to discover realities not through names but 
directly—is probably beyond our comprehension. 
As a final question, if the name-makers did act on a belief in universal 
and continuous flux and motion, were they right? Not if what Socrates 
' dreams' (and Cratylus says ' must be so') is correct, namely that there 
is an absolute beauty and absolute good, 'and so with all existing 
things'. It is these that demand our attention, not particular beautiful 
things and the question whether they are in flux, but beauty itself, 
which never changes. What is continually changing cannot be spoken 
of or known. We cannot say 'this' or 'such'; it is not anything, for if it 
stays the same for a moment, it is not changing. Nor can we know it, 
for even as the knower approaches it becomes something different. 
Knowledge itself cannot exist. Either it remains the same (which  
contradicts the flux doctrine) or if the very form of knowledge is always 
changing, it will no longer be knowledge. No, if there is always a 
knower and a known, if there is beauty, and goodness and every other 
existing thing, they bear no resemblance to flux or motion.2 
Whichever view is right, no man of sense will trust to names and 
their makers as proof that everything runs like a leaky pot or a cold in 
the nose. It may be so, but the question calls for more hard thought. 
Cratylus promises not to shirk this, but all his study so far has confirmed 
him in the Heraclitean view, and he hopes S. will think it over too. 
1 Lit. ' What is other than and different from them would signify not them but what is other 
and different.' The meaning and reference of this are puzzling. Does P. no longer believe that all 
reality is akin (Meno 81 c)? Or is it a hint that names do not after all resemble realities, as S. has 
all along said they do? But in the very next sentence he repeats this. 
3 See further on this passage pp. 8i f. below. 
15 
Cratylus 
Comment1 
This is real dialectic, with Plato at his most teasing in his effort to make 
us think. To examine the topical question of 'correctness of names' 
from all sides, he has taken full advantage of the dialogue form. It 
enables him to set out opposing theories, to show that neither is wholly 
right and conclude only that the matter needs more thought. Socrates 
is as wayward and wicked as he has ever been, taking first one side and 
then the other. No wonder scholars have differed widely over what 
Plato was trying to do, but in fact he leaves no doubt what was  
important to him. Throughout the discussions his own convictions (now 
familiar to us) flash in and out among those he thinks absurd, in the 
unique way which we know from his treatment of Sophists like Hippias, 
and are openly and plainly stated at the end. His choice of the status of 
names as subject had a double motive: first, it was a recent topic of 
debate among Sophists on which he could show up their errors with 
his favourite blend of seriousness and humour (especially the latter), 
and secondly it affected his cherished doctrine of Forms. (Proclus, In 
Crat^ p. 3 Pasqu., remarked that to understand the correctness of names 
was a necessary preliminary to dialectic.) 
The idea that the correct concept of a thing must be inherent in its 
name was not confined to professionals but is attested by many passages 
in Greek literature, especially tragedy.2 Since it is foreign to our 
thought, Wilamowitz did a service in pointing out what a natural, 
indeed inevitable assumption it was at the time {PL I, 287 f.), and we 
need not be surprised at the Heraclitean Cratylus accepting it, even if 
we suspect that it is Socrates who puts into his head the welcome 
thought that any etymology can be made to support the flux-doctrine. 
1 I gladly acknowledge that in addition to published sources I have received help from a 
notebook of Cornford's containing notes on the Crat. References to Cornford in what follows 
are to these notes. 
3 Kirk (HCF 119) mentioned some examples from tragedy, which could easily be multiplied. 
A good one is Eur. Tro. 889 f., connecting Aphrodite with ?????, ' foolish', instead of the more 
usual ?????, 'foam', which S. retains at 406d, though calling it her 'playful' name. There is 
evidence that Democritus, who certainly theorized on language (vol. in, 474-6), etymologized 
the names of gods (fr. 2). Kahn has pointed out that this too, and the emphasis on names and 
etymology in the fourth-cent. Orphic papyrus from Derveni, illustrate an extant fashion of 
explaining divine names allegorically which, if we knew more about it, might throw light on S.'s 
behaviour under the influence of Euthyphro. See esp. Crat. 401 ff. and Kahn in Exegesis 15 5 f. 
16 
Comment 
Conversation with Hermogenes. In his discussion with Hermogenes, 
Socrates constructs a theory, such as Cratylus (who has so far refused 
explanation) might bring against him, of how names can have natural 
rightness. Hence he gives it a Heraclitean basis as agreeable to Cratylus, 
who in fact declares it all much to his mind D28 c). As to Hermogenes's 
extreme view that even a single individual is entitled to call a man 
'horse',1 though all the world call him 'man', Socrates himself in the 
Charmides gave Critias licence to use any name he chose, provided he 
made clear what he meant by it. Grote accused Hermogenes of  
contradicting himself, because convention and agreement imply an  
intention to serve communication, which they would not do if everyone had 
his private vocabulary. But Hermogenes has not mentioned  
communication, and at 388b is unable to say what a name is for until Socrates 
prompts him. His present point is different, that the sounds in 'horse' 
can just as suitably indicate a man as a horse. If others acquiesced in the 
change, nothing would be lost, for there is no natural affinity, as 
Cratylus would claim, between a thing and its name.2 
A fallacy of division? At 385 b Socrates argues that just as statements can 
be true or false, so can the names of which they are composed.  
Robinson comments that the argument is bad,' for names have no truth value, 
and the reason given for saying that they do is a fallacy of division '.3 
He thus offers two objections, (a) Socrates's words imply the universal 
proposition that if a whole has a certain characteristic, so will its parts. 
This is obviously a fallacy, (b) Socrates is wrong in this particular case 
because names have no ' truth value'. Here we should note the definition 
1 Apparently he could 'agree with himself D35a); but in any case H. uses indifferently 
'agreement' (???????, ????????) and 'wont' (????). He is usually thought to be confusing two 
theories of names, the 'Humpty Dumpty' one that the name of anything is what I choose to call 
it, and the more serious view of language as a social institution with word-thing correlations 
conventionally established by the tradition of a particular language (Kahn, Exegesis i58f.). 
* Cf. B. Heath, JPh 17 A888), 195. 
3 Essays 123. S.'s statement had already been called a sophism by Steinthal in 1890, and 
defended by Goldschmidt, Essai 51 f. R. has been challenged by Lorenz and Mittelstrass in Mind 
1967 and Luce in CQ 1969. (One must of course avoid the petitio principii of replying that it is 
validated by the theory of natural names, which is what S. is using it to prove.) Aristotle too 
distinguished between names and propositions by saying of a name: '"Man" signifies something, 
but not that it exists or does not exist' (or 'is or is not the case', 2???? ? ??? j-???? De int. 
i6bi6~S). So P. says that one can point to a portrait and say 'man' or 'woman', thus conveying 
information or misinformation. 
3 
17 
GHG 
Cratylus 
of'true' at 385 b7 (remembering that alethes means 'real' or 'genuine' 
as opposed to 'imitation', as well as 'true' in describing statements: 
see p. 69 below): 'a true logos is one which speaks of (or describes) 
things as they are'. In Greek eyes names themselves (including proper 
names, nouns and adjectives) fulfilled this condition,1 and as we have 
noticed, most Greek names are transparently descriptive. Another 
example, quoted by Luce (I.e. 225), is Aesch. P. V. 85 f.:' Falsely do the 
gods call you Prometheus'—Prometheus meaning 'foresight'. For 
nouns and adjectives the same is obvious when they are compound like 
philosophos, and it was easy to believe that when simple they had the 
same function, even if it was now difficult to detect. This is what 
Socrates is going to try to demonstrate. It is also true that a single name 
or word can serve as a statement, as 'Cratylus' answers the question 
' Who is that?',' Walking' answers' What is he doing?' As a newspaper 
heading,' Parliament' gives information, true if above a report of  
proceedings in Parliament, false if above the report of a murder. It would 
be unfair to object that a word so used is virtually a statement, or implies 
a statement, e.g. 'Cratylus is walking' or 'What follows is a  
parliamentary report'. Only one word is uttere<I and that word gives the 
information. Plato is not speaking of words apart from any context, but 
as 'parts of discourse'.2 
Hermogenes and Protagoras C86a). Hermogenes need not have worried 
about his reluctance to accept Protagorean relativism,3 for his own 
theory of names by no means implies it. He holds that there exist 
objects with permanent and definable characteristics, e.g. (to take the 
1 A note of Cornford's is worth transcribing here: 'The use of ?????* is only intelligible by 
keeping its definition in mind, ?? ?? ???? ?????? ? ?????, (?) If the ?? in question is a relation 
between two things, the corresponding speech is a proposition affirming such a relation, (jb) If the 
?? is a thing, the corresponding speech is a name. Just as the proposition is true if it rightly 
reflects the objectively existing relation, so the name is true if it rightly reflects the objectively 
existing thing, i.e. speaks of it ? ?????. This it can do if, and only if, the material (sounds) has 
a natural (?????) correspondence with the forms of things (???(??) as is shown to be the case at 
434 a-b.' 
2 Cf. Luce's 1969 article passim, esp. 224 f.: ? name, by being uttered in a context, i.e. as a 
label for a person or thing, acquires truth-value.' For another treatment of Robinson's criticism 
see now Kahn, Exegesis 159-61, and just as this book was finished there has appeared Mary 
Richardson, 'True and false names in the " Cratylus" \ Phron. 1976. 
3 For Protagoras's relativism or subjectivism see vol. in, 183-8. His relation to Euthydemus 
(introduced here at 386d) is mentioned at 186 n. 1. 
18 
Comment 
example from Phdr. 260b) 'tame animal with the longest ears', but that 
it makes no difference whether you call it donkey or horse, or for that 
matter 'blip' or 'cump', since any name is only an arbitrarily chosen 
label. Protagoras's theory was that though language is constant, and 
two men refer to the same sensation when they say 'cold' or 'warm', 
one may may feel cold in the same situation in which another feels 
warm. They cannot contradict each other, not because one man has a 
private language in which 'cold' is the name for the sensation which 
others call 'warm', but because they are having different sensations. 
Protagoras is dismissed very briefly here, for Plato is after other game. 
He gets his turn in the Theaetetus. 
Essence and form. At 386d-e, Hermogenes agrees with Socrates that, if 
Protagoras was wrong, ' things must have a certain stable essence (or 
being, ousid) of their own, not in relation to us, dragged this way and 
that by our own imaginings, but as naturally constituted, in themselves 
and in relation to their own essence'. It will be convenient to take here 
some other passages in which the Platonic Socrates gives his own view 
on objective reality and its relation to names. 
423 e: * Don't you think each thing has an essence just as it has a 
colour and the other qualities we mentioned just now? Isn't there an 
essence of colour and sound themselves, and everything else that is 
rightly said to "be"?' 
393d: The precise syllables and letters do not matter 'so long as the 
essence of the object prevails and is revealed in the name'. 
422d: 'The correctness of the names we have reviewed meant that 
they revealed each object as it was.' Similarly 428c: 'We agree that the 
correctness of a name means that it shall show the object as it is.' 
438d: The discussion has now shown that names by their formation 
appear to give contradictory explanations of reality, so ' we must look 
for something else' to show us 'the true nature of existing things'. 
There is also the statement of the twofold function of names at 388b: 
they are instruments used for (a) informing, (b) distinguishing things 
as they really are. 
Finally there is the much discussed passage 386eff., where Socrates 
introduces the well-worn analogy from the crafts which made him so 
l9 
3-2 
Cratylus 
unpopular in real life. A name is an instrument used for a purpose, just 
like an auger or a shuttle, and it is the purpose which the shuttle-maker 
has in mind as he shapes his wood. The proper definition of a shuttle is 
not 'a piece of wood of such-and-such shape and dimensions', but 'a 
tool ideally fitted to separate the threads \ This therefore may be called 
'the shuttle itself or the eidos 'shuttle' (which to Hermogenes, a 
Socratic but not a Platonist, could mean either form, appearance, or 
class, species), and this is, or should be, the object of a definition, 'what 
a shuttle essentially is' C89b 5)—the old Socratic lesson as exemplified 
e.g. in Hippias Major and Meno. Moreover, since there are different 
shuttles for different materials and styles of weaving, the maker of an 
individual shuttle must give his wood both the general eidos of shuttle 
and the specific character (physis) required for its special work.1 
In the context of the dialogue the application to names is this. Steel 
(to borrow an illustration from Cornford) is used to inscribe stone, a 
diamond to inscribe glass. Fur each purpose the eidos of cutting tool is 
put into a different kind of matter, and, owing to the different quality of 
the material to be treated, the diamond is shaped into a point for 
scratching, the steel into a blade (chisel) for chopping. The general 
eidos C90a 5-6) or essence C89d6~7) of a name is to be an informative 
and diacritical instrument. The subordinate eidos, corresponding to 
sharp point or flat blade, appears in the difference between significant 
letters (r = motion etc.), each combination of which indicates a 
particular kind of things. The difference in matter (as between the steel 
of two chisels), which can be ignored, is the difference between (a) a 
Greek and a foreign combination of letters with the same meaning, or 
(b) two Greek combinations with the same meaning, like Hector and 
Astyanax, or iatros (physician) and akester (healer). At 390c the true 
craftsman in names is he who ' looks solely to that which is the natural 
name for each thing and can implant its form in the letters and syllables ',2 
but it has just been said that he can only do this by working under the 
1 This may be an indication of how Socrates thought that the relativity of the concept 'good' 
was compatible with the notion of a general eidos. See vol. in, 463 ff. 
* Aristotle (De int. 17a 1) said that what is ???? ???????? cannot be an ???????. His point 
seems to be the one made acutely by Crombie (EPD 11, 477), that 'whereas the function of a 
shuttle does determine its form, the function of a name does not'. Here P.'s whole argument, 
supported by the examples of' imitative' letters, is that it does, but by the end of the talk with 
Cratylus he has shown that his own view coincides with Aristotle's and Crombie's. 
20 
Comment 
instructions of the dialectician, that is, the philosopher. Plato is making 
the same point here as at Rep. 601 cff., where the user has knowledge 
and the maker only 'right opinion'. It is quite in his dialectical manner 
to make a looser statement first (the carpenter looks to the Form) and 
then refine on it.1 
Once again we see the teleological (that is, practical) basis of all 
Socraticism and Platonism.2 We also see how easily the Socratic view 
developed in Plato's mind into the belief in independent Forms existing 
prior to their material instantiations, for obviously the function of a 
shuttle preceded any actual shuttles; that is to say, no shuttles would 
have been made until someone had felt the need of an implement of 
that kind to do the work he wanted done. The Cratylus provides a 
clear example of the Form being spoken of as not only an internal 
character or essence but an ideal. It is what the weaver and the carpenter 
'look to' and embody in their material as far as they can. Plato adds 
that if a shuttle is damaged, the maker will not simply try to copy it but 
will consider ' that Form which he looked to when he made the broken 
one'. Strang3 quotes a fascinating modern parallel. The 'standard yard' 
was made in 1760. In 1834 it was damaged, 'and the commission set up 
to replace it decided to reconstruct it as accurately as possible in terms 
of its certified copies'. But in this they contravened an act of 1824 'that 
the restoration should be in terms of the length of a pendulum which, 
swinging in a vacuum in the latitude of London, should have a periodic 
time of two solar seconds exactly'. That was the ' Form', to which both 
maker and replacer had to 'look'. 
Plato might have been suspicious of even this formula, for a  
swinging pendulum suggests the world of sense, where accuracy is never 
more than approximate. (Cf. his treatment of astronomy, vol. iv, 524.) 
1 According to Ross (PTI 19), at 389C3-6 and 390b 1-2 P. speaks of a skilful carpenter as 
succeeding [completely] in embodying the Form in particulars. In the first passage he is pressing 
the words too hard and ignoring the point, which is only that the function of a tool demands that 
it be given an objective form or character, irrespective of our whims, if we want it to perform that 
function. You may find a hammer a more pleasing object than an axe, but if you want to chop 
wood you must subdue your aesthetic leanings and fit an axe-blade to the haft and not a  
hammerhead. In the second passage P. is actually saying that not the artificer but the user of his product 
knows 'the eidos proper to a shuttle'. 
* Socratic before it was Platonic. See vol. in, index s.v. Socrates: teleology. 
3 In Plato (ed. Vlastos) 1,188. He develops the theme as a fatal objection to P.'s 'paradigmat- 
ism\ 
21 
Cratylus 
In the Cratylus, as in the Phaedo, a Form (a) is both logically and 
temporally prior to its instantiations, (b) represents a perfection to 
which they can only approximate by imitating, or as he says elsewhere, 
'sharing in' them, (c) is an object of intelligence, not of sense.1 (For the 
last point see 423 d-e, where sensible colours and sounds are contrasted 
with the essence of colour and sound.) At the same time, the vocabulary 
of this doctrine has so much in common with the Socratic (as  
exemplified in the Euthyphro, vol. iv, ii4ff.) that Hermogenes can follow 
and approve the argument without (one may assume) an inkling of its 
transcendent implications. Of course the notion of an essential shuttle 
existing eternally, independent of the invention of weaving, is absurd,2 
and as we shall see, the range of application of the doctrine of Forms 
became a serious stumbling-block, of which Plato does not yet seem 
aware. At present it serves as an illustration, whose philosophical  
pitfalls have not yet struck him. His interest is not in weaving, but in the 
ethico-aesthetic Forms in which Socrates dealt—the Good, the  
Beautiful and their similars D39c). (Cf. vol. iv, 548-51.) 
If the function of names is to inform by distinguishing the essence of 
things, they become in fact potted definitions, and a definition is of 
necessity universal, a statement of the eidos^ which means both 
1 The question whether P.'s doctrine of Forms had advanced so far in Crat. has led to much 
discussion. An early sceptic was Ritter (N. Unters. 262ff., esp. 266). Ross (PTI19) says P. 'has 
not yet reached the point of thinking that an Idea is never perfectly exemplified, but only  
imitated', and Luce that the Crat. 'constitutes ... a stage distinctly prior to the position reached in 
the Phaedo* (Phron. 1965, 21: in a note he refers to other opinions). Contrast Hackforth (Phaedo 
9): 'Surely Plato could hardly have used plainer language to indicate that he conceives the Form 
as existing apart from its particulars, and indeed before any of its particulars.' He adds a criticism 
of Ross. Cf. however the strikingly similar language at Gorg. 503 d, and Dodds adloc. 
* The point brings out well the difference between the Platonic and Aristotelian conceptions 
of form. For Aristotle too the form is both logically and chronologically prior to the product, 
because it must pre-exist in the mind of the maker (Metaph. 1032332-^1). 
3 Or of the ????? (e.g. Laws 895 d). P.'s theory of definitions is in modern terms realist or 
essentialist, not, like most modern theories, nominal or linguistic (terms which mean that, in 
Mill's words (Logic, bk 1 ch. 8, 6), 'all definitions are of names, and of names only'). So Hospers, 
Phil. Anal. 54: 'In these cases [i.e. what are claimed as real definitions], what is defined is always 
a word or a phrase—a symbol. The language of "essence", however, may mislead us into  
thinking that we are defining things', and Russell and Whitehead, Pr. Math. 1, p. 11 (quoted by 
Abelson in his article'Definition', Ency. Phil. 2,319): ? definition is concerned wholly with the 
symbols, not with what they symbolize.' Moore on the other hand (Pr. Eth. 6-7) calls this kind 
of definition comparatively unimportant except in lexicography: the definitions he wants are 
' those which describe the real nature of the object or notion denoted by a word', only possible 
when the object or notion is something complex. Both Antisthenes and Plato had anticipated him 
in the latter point (see vol. m, 21 if. and Tht. 20id-202b). It led him to conclude that 'good' 
22 
Comment 
essential character (like ousid) and class. This is why Plato concentrates 
on common nouns rather than proper names. Even in the case of 
portraits he speaks (improbably) of mistaking a man's for a woman's, 
not (as would seem more natural) one of Hermogenes for one of 
Cratylus. In his etymologies he includes both, but once the proper 
name has been explained as descriptive of an essence, it immediately 
puts its holder in a class: there are many lords of cities besides Astyanax. 
It is the typos, general character or stamp, which the name must show 
D32e). The true power (dynamis) of names and their relation to 
realities, as Plato sees it, is made clear in the second part of the dialogue; 
but to say that they deal with essences is at least to this extent true, that 
only by using words can we exercise the uniquely human capacity to 
generalize (vol. iv, 427). 
The etymologies,} These are a bewildering and sometimes ludicrous  
collection. They reminded Crombie of Lewis Carroll, and certainly some 
are in the same class as the Gryphon's association of 'lesson' with 
'lessen'. Others are on sound lines and even correct, and the whole 
section is a regular encyclopaedia (as Goldschmidt called it) of Plato's 
knowledge of earlier and contemporary lore in physical science, 
cosmology, anthropology, and philosophical and religious ideas. The 
common element in the etymologies of 3910-427 d is that all are made 
to support the Heraclitean and Cratylean theory of cosmological flux. 
(See especially 411c.) No one knows how seriously to take them. One 
can hardly accept the correct or reasonable derivations as evidence 
when they are treated as exactly on a par with the silly ones. Grote, it is 
true, argued that in Plato's eyes none were extravagant, and cited 
examples of derivations apparently meant seriously elsewhere.2 Never- 
was indefinable, and P. to say that whereas dialectical argument was an indispensable preliminary 
to the grasp of its essence, the full comprehension of it could only be the result of an intuitive leap 
or sudden illumination. 
1 ' Greek etymologies are not to be compared with our scientific etymologies because they are a 
different thing; they must in fact be regarded as an attempt to penetrate the mystery of things: 
their meaning is philosophical, not linguistic.' (Untersteiner, Sophists 224 n. 42 ad fin.) 
* Phdr. 238c Spcos from ?>???, 244b ??????? = ?????? (very like Crat. in dismissing the ? as an 
'insensitive' addition) and 244c ?????????? from ??????, voOs and ??????? (though even Grote 
has doubts about this), Tim. 43c ???????? from ?????, 62a ?????? from ?????????. But 
Taylor (Comm. 269, 432) calls both Tim. examples 'fancies' or 'sportive', and it is incredible that 
those from Phdr. had any serious philological intent. At Tim. 45 b, P. offers for ????? the deriva- 
23 
Cratylus 
theless—to cite one example only—I do not believe Plato thought the 
name of Kronos to be derived from the clean-swept purity of his mind 
C96b). 
Others have maintained that although individual etymologies may 
be bad, they illustrate sound principles scarcely out of date even today.1 
Such are the effect of euphony D04 d) and other changes in use, and the 
necessity to seek out the oldest form D18 e), the beginnings of  
comparative philology (especially at 410a), 'sound-symbolism' and the clear 
way in which the theory of speech as vocal gesture is distinguished 
from the cruder onomatopoeia. Yet the principles themselves mix sense 
and nonsense. Women, for instance, are said to be, or to have been, 
particularly fond of the sound of/ and </, and to be especially  
conservative in their speech Di8b-c). We may note too that Socrates himself 
finally rejects the resort to foreign origin as 'escapist' D26a), and about 
sound-symbolism (more accurately imitation of essences through 
movements and positions of the organs of speech), which he admits 
seems ridiculous, he does not say that it is a true explanation but that it 
is the only rational way to defend the correctness of primary names 
(which in fact neither Hermogenes nor Plato himself believes in). Its 
uselessness is subtly shown in the very course of expounding it, e.g. by 
the true observation that even names with an obvious meaning may 
have been given for the wrong reasons (family connexions, wishful 
thinking) and have therefore no natural affinity with their nominates 
C94d-e, 397b), and that some names have been twisted to unintel- 
ligibility by 'people with no regard for the truth' until 'any name can 
be fitted to any object' Di4C-d). Again, the professed purpose of the 
exercise is to show that words are connected with their objects by a 
theory of significant sounds, yet after refuting Cratylus on the basis of 
his assumption that they are so connected, Socrates upsets the whole 
thing by showing that etymologies could as easily be made to support a 
tion from ?????? which at Crat. 418 d he explicitly rejects. Other Platonic etymologies are given 
by Meridier on p. 18, n. 2. More impressive is the sober Aristotle: HA 493322 ????? from 
???????, EN ??^??-} ??????? from ????, Phys. 198b22 ????????? from ???? ?????. Grote adds 
??????? from ???? ?? ?????, but fr. 102 Rose suggests that this was not Aristotle's own idea. 
G. also quotes opinions from later antiquity to support the view that P. was serious. His 2nd vol., 
pp. 518-29 with their notes, should not be missed, if only as a brilliant display of scholarship. 
1 See e.g. Jowett, Dials, m, 2, Friedlander, PL 11, 206 with n. 30, PfeifTer, Hist. CI. Sch. 1, 63 f., 
Meridier, Crat. 26. 
24 
Comment 
theory of immobility as one of flux, even offering contrary explanations 
of the same word.1 
Finally, whether or not either single examples or general principles 
are defensible, the idea that Plato meant them as a serious linguistic 
exercise is ruled out by the consistently humorous and ironic vein in 
which they are proposed. This I have tried to bring out in the summary. 
One of the things which has to be accepted about Plato and makes him 
so baffling a philosopher, and so delightful a writer, is the way that his 
Socrates plays with a naive or Sophistic respondent by tumbling  
together with an equally straight face absurdities and deeply held  
convictions. Perhaps agreement will never be reached, but for one reader at 
least the philological lessons are in the former class, a take-off of the 
current pseudo-science of etymology. 
The right relation between names and reality. It may be that the  
terminology in 386e-88e closely resembles that at Rep. 596aff.,2 which 
ostensibly posits a single form {eidos) for 'every set of things to which 
we apply the same name'. This would conflict with Crat. 387b-d, 
where Socrates points out to Hermogenes that 'if a man speaks as 
things are intended by nature to be spoken of, and with the appropriate 
instrument [sc. name], his action—that is, his speech—will accomplish 
something. Otherwise he will be in error, and his action nullified.' 
I hope I have shown, however, that the meaning of the Republic 
passage is not quite what it appears to be on the surface. (See vol. iv, 
5 50.) The purpose of names is to classify according to essence C88 b-c), 
but a wrong name may be given, and it will then be untrue that  
everything included under it will have the same eidos. If names are entirely 
conventional, there is no guarantee that a common name indicates a 
common idea. 
The Socratic conception of eidos and method of definition left a 
series of problems for Plato, of which he only gradually became aware, 
when the instability of particulars forced him to conclude that this 
1 ????????. See 412a and 437a. Note also that on the theory of sound-symbolism names (in 
spite of 423 e) can only imitate sensible attributes like speed, smoothness and shape, whereas 
essences or Forms are for Plato ?????, grasped by the intellect not the senses. And ????? are the 
only true objects of knowledge, therefore knowledge cannot be obtained from words. 
3 So Allan, AJP 1954, 281. 
25 
Cratylus 
'Form' was not only the general character determining a class but also 
a separate, perfect exemplar to which the particulars could only partially 
aspire. The Socratic method of induction from cases where the same 
term is currently employed (justice, courage, self-control are all virtues, 
therefore they must share a common characteristic 'by which' they are 
virtues) could be taken to assume that use of the same name points to a 
resemblance between things, i.e. it infers from names to things. It looks 
as if the Cratylus is designed to correct this misconception. In its 
extreme form the assumption is that of Cratylus: 'It is quite simple: 
whoever knows the names knows the things too' D35d). Against this 
* Socrates' argues: (a) Whoever bestowed names doubtless formed 
them according to his notion of the things, but he could have been 
mistaken D36a-b);1 (b) Knowledge of things cannot depend solely on 
knowledge of names, for there must have been a first name-maker, and 
how did he get his knowledge? Things then must be knowable,  
somehow, directly ('through themselves and each other') D38c),2 and the 
correctness of names, as he said to Hermogenes, must be tested by their 
capacity to distinguish the essences, or indwelling nature, of things 
D22 d). 
Cornford put forward a tempting possibility, in terms of the earlier 
and later theory of Forms. On this view the assumption that a common 
name always indicates a common eidos would be both a Socratic and an 
early Platonic one, which the Cratylus is designed to examine. It shows 
that the assumption is unjustifiable and that consequently the basis of 
the early theory must be modified to render it independent of the 
caprices and imperfections of language. 'Is not', he continued, 'this 
modification the chief difference between the earlier and later theories, 
the later being based not on common names or on language at all, but 
on natural kinds distinguished by observed characteristics?' One 
thinks of Pol. 262 b-e, with its warning not to think you are dividing 
'according to form' because something is given a single name, or Tim. 
83 c which speaks of 'someone who can look at a variety of things and 
see within them a single genus justifying one name for all'. 
1 He might, says S., have been wrong in his initial assumption; and the errors would multiply 
as he forced the rest into agreement with it, as occurs in geometrical proofs. The importance of 
testing a hypothesis in every way possible is brought out in Phaedo. See vol. iv, 352f. 
* Haag p. 5 compares Tht. 186 a-b. 
16 
Comment 
Nevertheless this would be difficult to maintain. As an illustration of 
the earlier theory Cornford cited Rep. 596 a about positing a Form for 
everything with the same name. But in the same dialogue D54a) 
Socrates distinguishes eristics, or sharp debaters, from dialectical 
philosophers as people 'who cannot analyse their subject into its 
natural kinds (???????????? ???' ????) but chase a contradiction in it 
through going by the mere name'.1 The need to 'divide by (natural) 
kinds' has met us in the Phaedrus (vol. iv, 428) and recurs here, when 
at 424c-d Socrates gives Hermogenes a lesson in the method by  
showing how letters can be divided into natural kinds (????????? ???' ????) 
which can be fitted to the natural divisions between things. It goes back 
to Socrates, and perhaps further. A passage from the Hippocratic De 
arte 2 is worth quoting: ? believe the technai took even their names 
from their natures (eide\ for it would be absurd and impossible to 
suppose that the forms originate from names. Names are conventions 
imposed on nature, but forms are not conventions but natural growths.' 
As for Socrates, according to Xenophon he suggested that dialectic was 
so called because its practitioners in their converse 'divided things 
according to their classes' (??????????? ???? ????).2 Evidently the 
method of definition by division, exemplified in the Sophist and 
Statesman^ was not a new departure, but a technical elaboration of 
something with which Plato was familiar from the beginning. It and 
'collection' were never, even in their earliest stages, dependent on 
inference from names to things. The assumption implied in the Socratic 
method was rather' that the kinds or classes to which particulars belong, 
the "forms" which they possess, have a quasi-substantial nature and 
hence a stability which enables the essence of each to be grasped, 
1 The context is S.'s proposal that women and men should share the same occupations. To 
object that they are 'different' could be like saying that bald men and men with hair should not 
pursue the same trade. The question is, in what respect are they different? The mere name 
'different' is no help in determining the natures of men and women. 
% Xen. Mem. 4.5.12. (????? and ????? are used indifferently in the passage on ???????? at 
Pol. 262d-e.) See also vol. m, 440 f. for S., and for the fifth cent, in general the ref. to Morrison 
on p. 204. Note in contrast to the Socratic ???' ???? ???????? that Prodicus is more than once 
ironically lauded for his skill at ??????? ???????? {Charm. 163d, Laches 19yd). De arte is 
arguing against those who would deny reality to the ??????, especially medicine. I cannot 
believe that it is 'after' Plato, as some have suggested, rather than belonging to the fifth-cent, 
controversy over the status of names. (Heinimann, N.u.Ph. 160 dates it thus.) My translation of 
?? ??????? ?????? in vol. in, p. 204 probably needs correcting. 
27 
Cratylus 
described, and clearly distinguished from all other essences'.1 As the 
Phaedo says A02 b),' Things take their names from the Forms in which 
they participate.' 
But modern comment on Rep. 596 a shows how easily it could all be 
misunderstood as making correct classification and definition depend 
on the accidents of nomenclature. Such misapprehension would have 
horrified Plato, as undoing his master's work and striking at the very 
roots of the Socratic as well as his own belief in Forms, and the Cratylus 
may be seen, not as a correction of his own earlier teaching but a 
defence of it against unjustified criticism, and perhaps a lesson for his 
pupils.2 
The upshot of the Cratylus is that names do give information by 
distinguishing between classes or essences of things ('It wasn't a 
burglar, only a cat'), but only if the essences are known beforehand 
D38 a-b). We are left therefore with the problem of how we can know 
and distinguish between things in themselves. They must be known 
1 Vol. in, 440. There is an argument about this which goes well back into the last century, and 
in which prejudice has played a considerable part, as Stenzel rightly saw. Reacting against it, he 
declared (PMD 80) that Rep. 454a in its context has nothing to do with the technical procedure 
of the division of a genus in Soph. P., he says, used the word ???????? even in the earlier dialogues 
to mean * division into parts' (what else could it mean?), but it must have been quite different 
considerations that led him to the consciously-held theory oiSoph. and Pol. Cornford too (PTK 
i84f.) contrasts Socratic and Platonic methods in this respect, taking Meno as typical of the 
former, and Dodds (Gorg. 226) calls diairesis a Platonic invention, though seeing it already (as one 
must) in Gorg. Soph, and Pol. certainly represent a development in technique, but I believe it is 
going too far to say that this has nothing to do with Socratic method and arose from quite 
different considerations. (Cf. esp. vol. iv, 430.) Cornford {PTK 180) says that Collection must 
not be confused with the Socratic muster of individual instances', because it is confined to Forms, 
but S. himself (if, as Cornford does, we take P.'s earlier dialogues to illustrate his method) 
normally operated with forms or species as units. It is the main thesis of Sayre's ? AM that there 
is no sharp distinction between the method of hypothesis in Phaedo or Rep. and the collection 
and division of the later dialogues: they are a single method in two stages of development. Ten 
years after the German original of PMD, Stenzel himself wrote (RE, 2. Reihe, 5. Halbb. 862): 
'P. saw in the general notion of "separating", sharpened perhaps by S.,... an earlier form of his 
technical method of definition by diairesis.' 
* Wilamowitz (PI. 1, 289) argued that P. himself had once been very attracted to the idea that 
the essence of things could be found in words, and the Crat. was written to rescue himself and his 
pupils from this illusion. He declares it wrong, but enjoys playing with it as only a man could who 
had ventured far on this path before discovering that it led nowhere. (One might think of S.'s 
reluctance to give up the imitation' theory of words at 535 c. Cf. also the evidence presented with 
admirable impartiality by Goldschmidt, Essai 185-99.) This is possible, and would explain very 
satisfactorily the length and gusto of the etymological section, which some scholars have found 
disproportionate, though it is its own justification, far too ingenious and entertaining to merit 
such strictures as Nferidier's (p. 33): * P. croyait avoir ses raisons, mais il est certain que du point 
de vue artistique l'dconomie de l'ceuvre en a souffert.' 
28 
Comment 
directly, 'through themselves and each other' D38c), but how to do 
this is set aside as a problem ' too big for you and me'. It is reserved for 
the Theaetetus and Sophist. The obvious way would seem to be through 
sensation, whose claim to be called knowledge is fully discussed in the 
former. In the Sophist the hint that there is a definite relationship 
between the Forms ('through each other') is taken up and developed. 
From our knowledge of other dialogues, especially Phaedrus and 
Phaedo, we can see the answer which, so far at least, has appealed to 
Plato himself. Sensation must indeed be the starting-point (Pkdr. 249 b, 
Phaedo 74 b). From sensations1 all human beings have the power of 
forming general concepts, making possible the use of general terms. 
The philosopher however, by his skill in dialectic, carries further the 
process of recollecting the perfect Forms which his soul saw when free 
from the body, until, having recovered them, he can use them as 
standards for his classification of things on earth. The effective  
resemblance is not between things and their names, but between things 
and Forms, those steadfast, unchanging Forms which Socrates  
suddenly brings in at the end and which Cratylus cannot deny—not, 
surely, because he held any Platonic theory of their transcendence, but 
as Protagoras did, because it seems absurd to say that there is no such 
thing as beauty or goodness. (Cf. vol. iv, 223.) 
What is meant by correctness of names? This does not settle one of the 
central questions raised by the Cratylus: how serious is Plato in saying 
that names reveal the nature of their objects by actually imitating them 
in sound? It is not decided by saying that the original legislator on 
names worked with a knowledge of the things named. Socrates's own 
purpose (which as we have often reminded ourselves was ultimately 
moral or social) did not require any resemblance between words and 
things but only consistency. Assuming as he did that 'justice itself 
existed and was immutable, it did not matter whether within a society 
it was called justice, dikaiosyne or Gerechtigkeit, provided its users 
indicated its true nature when they used the same word, having 
through the work of the dialectician rid themselves of confusion such 
1 Perhaps better perceptions'. P. was not worried by any distinction between these and pure 
sense-data. 
29 
Cratylus 
that one meant * obedience to the laws' and another * the right of the 
stronger'.1 
Roth has pointed out2 that not only are there two theories of naturally 
correct names in the Cratylus, but also, though not formally inconsistent 
themselves, they rest on inconsistent assumptions. The proper  
conclusion of the first is that a name is only correct if it makes clear the nature 
of the things it names; of the second, that a name is only correct if its 
letters and syllables imitate the nature of the thing named. In drawing 
the first conclusion Socrates insists that the letters and syllables chosen 
have no bearing on it whatever. At 390a the legislator on names does 
his work well provided he renders the form proper to each thing 'in 
whatever syllables, here or elsewhere'; and at 394b the expert in names 
in considering their dynameis will not be disturbed by a few changes or 
even if 'the force of a name is expressed in entirely different letters'. 
These are only the material, and the same form can be realized in 
different materials C89d-e). The validity of the second theory, on the 
other hand, obviously depends on the assumption that correctness of 
name is equivalent to correctness of vocal sound. At 433d~34a Socrates 
demands that Cratylus choose between the convention and resemblance 
theories as the only alternatives, and Cratylus naturally prefers the 
latter, which Socrates then (a) disputes by showing that even if some 
words indicate their objects by resembling them, others can do so by 
other means, and (b) makes fun of by showing that it could support a 
static theory of the universe as easily as Cratylus's own theory of flux. 
So we reach the seriously meant conclusion of the whole dialogue, that 
names offer no help in discovering the essential natures of things, 
though they serve to communicate those natures when known. With 
cats and burglars this is easy, but not so with the supremely important 
Forms of Good, Beautiful and Just, and others which are the Platonic 
philosopher's primary concern.3 
1 This point is made in the Crat. itself at 4346-353. 
* Diss. 1969, p. 88. (See p. 5 n. 3 above.) 
3 Ale. /, iiib-na is relevant here. 
30 
Comment 
ADDITIONAL NOTE: AN IDEAL LANGUAGE? 
Since Benfey's monograph in 1866, a question that has interested many is 
whether in the Cratylus Plato had in mind as an aim the creation of an ideal, 
artificial or technical language. This has been inferred from the analogy with 
the crafts, especially the comparison between carpenter and legislator and 
their relation to their respective overseers C90 b-d). Thus Grote wrote with 
reference to this passage {PL 11, 506): 'Plato aspires here to a philosophical 
language fit for those who conversed with forms or essences: something like 
... a technical nomenclature.' And in modern times Weingartner {Unity 35): 
' It seems that Plato was looking forward to a technical language which could 
reflect the classifications that result when dialectic becomes collection and 
division.' Runciman {PLE 21, n. 4 to p. 20) says that Grote was probably 
right in his view 'that Plato thought an absolute standard of naming to be 
theoretically desirable though not existing in fact', and supports this by 
reference to three other dialogues. I do not wish to continue the debate 
beyond what can be gathered from the above comments. The question is 
only peripheral to the main aims of the dialogue, and statements like those of 
438c and 439a tell strongly against a positive answer: names are only  
confusing, so that we must find something other than names to reveal the truth; 
and even if we could learn from them, there is a better way. If one must 
answer for Plato, therefore, it will be in the negative, but in a dialogue in 
which, on everything but the main point, he so skilfully covers up his tracks, 
it would be rash to dogmatize.1 
The Seventh Letter C43a) states that the names 'round' and 'straight' 
have no permanent validity, and could as well be reversed. They would be 
just as well established for those who changed them round. This is often 
quoted in connexion with the Cratylus, but it is dangerous to wrench it thus 
from its context, where it is only ancillary to a particular point, that nothing 
in the sensible world contains a pure quality unmixed with its opposite. No 
sensible circle, whether drawn or turned on a lathe, is perfectly circular. All 
contain elements of straightness as well. 
1 See also the judicious remarks of Goldschmidt, Essai 199-206. Since the above was written, 
Kahn has appeared on the same side {Exegesis 167). As evidence for * a vision of an ideal language* 
Anagnostopoulos quotes 424c-2$a. See his article on 'The significance of P.'s Crau * in R. of 
Metaph. 1973-4, 327. 
3» 
II 
PARMENIDES, THEAETETUS, 
SOPHIST, POLITICUS 
INTRODUCTION 
With these dialogues Plato's thought takes a remarkable turn. So far 
the figure of Socrates has dominated the rest and his point of view has 
prevailed. His companions are either convinced or at least silenced, and 
the reader is plainly meant to follow his lead. Secondly, the assumption, 
or hypothesis, of the existence of unchanging Forms—the Good, the 
Beautiful and the rest—which though separate from particular actions 
or things are in some way responsible for their being what they are, has 
never been challenged. It is received as something well known and 
accepted, and used as the basic premise from which deductions can 
safely be made. (Cf. especially Phaedo ioob-c, Rep. 476a, Crat. 439 c.) 
Suddenly in the Parmenides we meet a new Socrates, a very young 
man, unsure of himself and putting forward this same hypothesis, of 
separate Forms and particulars that * share in' them, only to have it 
attacked by the old and famous Parmenides, who counters all his  
arguments in its defence and after leaving him helpless kindly offers to give 
him the lesson in method which he so obviously needs. For the first 
time, what we have come to regard as the corner-stone of Platonism, 
bound up with visions of an immortal soul and a place beyond the 
heavens, is itself made the subject of a searching examination. In the 
Theaetetus Socrates is again in the lead, but an attempt is made to define 
knowledge without recourse to the Forms, whose existence is scarcely 
hinted at in the main argument. As to Sophist and Politicus* Socrates 
is reduced to a silent listener, and the discussion is conducted by a  
nameless character tailor-made for the occasion, and described precisely as a 
native of Elea and follower of Parmenides and Zeno, but nevertheless 
1 This latinized form of the Greek word for Statesman' (abbr. Pol.) is usual in England, 
though in some countries Pol. is used for the Republic (Politeia). 
32 
Introduction 
no eristic or logic-chopper and not afraid of challenging his revered 
teacher himself. In the Sophist he severely criticizes some people 
referred to as 'friends of Forms' for their belief that reality is  
changeless. There and in the Statesman the technique of defining by 'collection 
and division', explained in the Phaedrtis (vol. iv, 427-31), is refined and 
developed into the primary method of reaching a definition. 
Plato makes clear that he meant the four dialogues to be read in  
conjunction, and in the order given above.1 In both Theaetetus A83 c) and 
Sophist B17 c) Socrates mentions his long-past meeting with the aged 
Parmenides. In the Sophist the three speakers of the Theaetetus meet 
again 'according to yesterday's agreement' and introduce the visitor 
who is asked to explain the nature and mutual relations of three types: 
Sophist, Statesman and Philosopher; and the Statesman begins with 
explicit references to the Sophist, and includes others at 258b, 266d, 
284b and 286b. Theaetetus talks to Socrates in the Theaetetus, to the 
visitor in the Sophist, and is present but 'let off' in the Statesman, 
where his place is taken by the younger Socrates, who has been 
silently present at the two earlier discussions.2 
(i) PARMENIDES 
The Parmenides, especially its second part, has had the strangest fate 
of any of Plato's dialogues. That he was a theist, deeply religious and 
with more than a touch of mysticism in him, no one would deny; nor 
would anyone be surprised at finding the Phaedo, Phaedrus or Timaeus 
cited in evidence of this. But that the dry antithetical arguments of the 
Parmenides about the One, sophistic in form at least and inseparable, 
one would have thought, from fifth-fourth-century controversy, 
should have been seen as an exposition of the sublimest truths of 
theology, is surely one of the oddest turns in the history of human 
1 This is probably also the order of composition, but since they were thought out as a group, 
it is of no great importance. For their relation to Phdr. see vol. iv, 396. Campbell, it is true {Tht. 
\v\ gives his reasons for supposing that 'It does not appear that at the time of writing the 
Theaetetus Plato had distinctly planned the other three', and McDowell disputes the priority of 
Parm. to Tht. See his Tht. p. 113 and the other notes there referred to. Mrs Walker (PR 1938, 
503) writes of * the advance of the Parmenides beyond the Philebus\ but this is a highly unusual 
view. 
2 For his presence in Tht. see i47d. 
4 
33 
GHG 
ParmenideSy Theaetetus, Sophist^ Politicus 
thought. Yet the Neoplatonists claimed to see in the One their own 
highest, ineffable and unknowable God, and as such it passed into 
medieval and later Christianity and into philosophy as far as Hegel. 
Even the analytic approach of the present century has its rivals, as in 
WahPs talk A926) of a union of * transcendent mysticism and immanent 
pantheism', and Wundt's conclusion A935) that 'the Neoplatonists 
were not so far from Plato's doctrine as is often believed today'. 
Today's disputes go a long way back, for Proclus himself divided 
earlier interpreters into a logical and a metaphysical school.1 
Date. Of its position in the series of dialogues enough has already been 
said. It has been conjectured (Ritter, Essence 28) that it was written 
during Plato's second sojourn in Syracuse after Dion's exile, and if this 
is hardly susceptible of proof (cf. Taylor, introd. p. 2), the period 
370-367 is generally agreed to be very probable. It has also been 
asserted that the two parts into which the dialogue naturally falls were 
written independently at a considerable interval of time, and later 
stitched together. This is an old theory, revived by Ryle, if turning it 
inside out can be called revival, for according to the older view the 
second part was the earlier, whereas Ryle puts it later than the first.2 
The strongest argument is the complete change of style at 137 c from 
narrated dialogue, naming the speakers and with occasional mention of 
laughter and other descriptive touches, to direct speech and the  
abandonment of all pretence at narration. It is not however decisive, and the 
denial of an original and organic connexion between the two parts has 
not found general favour. 
Dramatic date. Plato is at pains to place the main discussion in its 
temporal setting. Parmenides, grey-haired and distinguished-looking, 
1 Wahl, Etude 43 and 88, Wundt, P.*s Parm. 26. Wyller's work A959 and later) has been 
described as 'a strange mixture of Proclus and Heidegger'. See Tigerstedt, Interpreting P. 143-7. 
Similar views are taken by Speiser A937) and Huber A951). For a brief historical summary of 
interpretations see Cornford, P. and P. v-ix, and for the Neoplatonists app. ? to Taylor's trans, 
and Wundt, o.c. 7-26. A more recent Neoplatonic interpretation of the Parm. is Hager's in 
Der Geist und das Eine A970). 
a Ryle in SPM 145, P.*s P. 287-93, correcting his earlier statement in SPM (p. 100) that 
'there is a clear connexion between the two parts'. His argument for separation is rebutted by 
Crombie, PR 1969, 372. For the earlier view of Apelt, Wilamowitz and Wundt see Wundt, 
P.*s Parm. 4-7. 
34 
Parmenides 
is about sixty-five, Zeno approaching forty, and Socrates * very young', 
so their meeting must have taken place about 450. But this is now a long 
time ago, nor do we get the story at first hand. As presented, the whole 
dialogue is narrated by one Cephalus of Clazomenae1 (otherwise  
unknown), who tells how he brought some fellow-philosophers to Athens 
to hear an account of the meeting from Plato's half-brother Antiphon 
who had it from Zeno's friend Pythodorus who was present. If 
Cephalus had to get it in this roundabout way, one can only assume 
(with Taylor) that the Athenian participants—Socrates, Pythodorus 
and Aristoteles—were dead, and Antiphon's narration took place after 
400. In the nearest comparable work, the Symposium, the narration 
takes place a mere sixteen years after the event, and the narrator has 
confirmed some points from Socrates himself. 
Setting and characters. The involved introduction may indicate that 
Plato wanted to prepare a reader for the fictional nature of the main 
discussion, but it could equally well be intended to emphasize its 
importance, if a group of philosophers from the home of Anaxagoras 
thought it worth while to journey to Athens to hear it over fifty years 
after the event. In any case, as always, Plato enjoys the personal 
touches for which it gives an opportunity. Cephalus and his friends 
meet Glaucon and Adeimantus in the market-place, who confirm that 
their half-brother Antiphon heard the discussion in his youth and took 
great pains to learn it by heart, though now his main interest is in 
horses. Together they go to Antiphon's house, and when he has 
settled the important matter of a new bit with his harness-maker, and 
grumbled about the difficulty of recalling the story now, he agrees to 
tell it. 
Zeno and Parmenides, on a visit to Athens, were staying with 
Pythodorus, at whose house in the Cerameicus the discussion took 
place. He is mentioned again in Ale.12s a paying pupil of Zeno,2 and as 
an Athenian general his name occurs several times in Thucydides in 
connexion with events in Sicily during the Peloponnesian War and the 
signing of the Peace of Nicias in 421. Also present, besides Socrates, 
1 The home of Anaxagoras. On the significance of this see Schofield in Mus. Helv. 1973, 4. 
a But on the payment see Vlastos in JHS 1975, 155-61. 
35 
4-2 
Parmenides, Theaetetus, Sophist, Politicus 
was Aristoteles, the youngest of the company, the Greek form of 
whose name will serve to distinguish him from the philosopher 
Aristotle. The unlikelihood of the idea1 that by the name Plato intended 
to link his famous pupil and critic with some objections to the doctrine 
of Forms is shown by more than one consideration: (a) Plato takes the 
trouble to point out at 127 d that he was the Aristoteles who became one 
of the Thirty Tyrants (known also from Xenophon, Hell. 2.3.2,13 and 
46); (b) he was pretty certainly writing at about the time when Aristotle 
first came down from Macedonia, aged about 17, to join the Academy. 
Of the historical Parmenides and Zeno nothing need be added here,2 
but it may be noted that the former is not a mere lay-figure. True, he 
does not always speak in the terms of his own Way of Truth. How 
could he? He is the first in a long line of philosophers who have  
propounded theories which they must ignore in practice. Could a solipsist 
act as such in communicating his ideas? But he suggests as subject for a 
dialectical exercise 'my own hypothesis about the One itself, and the 
main strength of his arguments in the first part lies in his historical 
denial of any possible connexion between the sensible and intelligible 
worlds, precisely the dilemma which Plato's doctrine of Forms was 
designed to solve. 3 
Part One A260-35d) 
Introductory conversation: Zenos arguments countered by the doctrine of 
Forms (i2jd-3oa). Zeno has been reading his treatise, the object of 
which, as he agrees with Socrates, was to defend Parmenides's thesis 
that 'All is One' A28a) indirectly by demonstrating that if there is a 
plurality things must have contrary characters, being e.g. both like and 
unlike, and this is absurd. Socrates counters with his familiar question: 
Are there not Forms of Similarity and Dissimilarity apart from particu- 
1 Revived by Wundt (o.c. 5 n. 2) and several other scholars, including recently Koutsouyanno- 
poulou in Platon 1966 and Newiger, Gorgias uber das Nichtseiende A973), 108. Against it see 
Taylor, trans. 1291*. and Cornford, P. and P. 109 n. 1. Bury's suggestion (/. Phil. 1894, 1761*.) 
that Aristotle could have already been pressing the objections because he had only taken them 
over from the Megarians is more ingenious than convincing. 
2 On Zeno and his relationship with Parmenides see Vlastos's article mentioned on previous 
page, n. 2. 
3 Schofield in CQ 1973, 44, sees P. as in the second part making Parmenides draw from his own 
hypothesis, and by Eleatic arguments, conclusions embarrassing to an Eleatic. 
36 
Parmenides 
lar people and things, which are similar or dissimilar according as they 
partake in one or the other? There is no reason why a particular should 
not partake in contrary Forms, as Socrates for instance is one person 
but comprises many parts. It would be surprising if Forms themselves 
could exhibit contrary characteristics by combining with their opposites 
—Similarity with Dissimilarity, Unity with Plurality, Rest with Motion 
and so on1—but the difficulty raised by Zeno and Parmenides affects 
only sensibles, not the intelligible Forms which exist apart. 
Zeno takes no further part in the discussion, which is conducted 
entirely by Parmenides. This is in keeping with Plato's opinion of the 
two men. Zeno is dismissed in the Phaedrus as a living demonstration 
that captious and contentious argument is not confined to lawyers and 
politicians. For Parmenides, who had changed the whole face of Greek 
philosophy, Plato had enormous respect tempered with fundamental 
disagreement.2 
The theory of Forms, as stated here and elaborated in reply to 
Parmenides's questions, is exactly that of the Phaedo^ so before  
encountering Parmenides we may recall its chief features. Forms (i) exist 
apart from particulars, as changeless and eternal exemplars, accessible 
to the mind in thought, but not to the senses. At the same time B) they 
are the causes of particulars being what they are, though one cannot be 
dogmatic about the relationship: particulars may be said to * share in' 
Forms or resemble them imperfectly G4c), or Forms can 'be present 
in' or 'associate with' particulars (iood). C) Hence one may distinguish 
between a Form in and by itself and its instantiation in a changing and 
perishable particular A02 d). D) There is a hint also of perfect instances 
of Forms, besides the Forms in physical beings, only mentioned in 
1 This is plainly what S. means, though he has been thought to be denying any combination 
of Forms, ?? ??????? are pairs of incompatible Forms like Unity and Plurality. Cf. Hicken in 
SPM 191. 
2 For Z. see Phdr. 261 d-e and Cornford, P. and P. 67f. (Note S.'s little joke at 128b: Z. 
denies the existence of many, and brings forward many arguments to prove his case.) For 
Parmenides, note S.'s expressions at Tht. 183 ? and Soph. 217 c modified by his fears in Soph. 
B16b) that an Eleatic will be a 'god of refutation', and the Eleatic's own criticisms of his master 
at 241 d and 242c. (Vlastos in JHS 1975, 150-5, sees rather differently the implications of Phdr. 
261 d-e.) 
3 The latest of many theses about the apparently severe criticism of the Forms in the Parm. is 
Zekl's of 1975. He holds that what is offered for criticism is not the genuine theory but a 'pale 
copy' of it, the immature effort of a not very bright pupil. The criticisms therefore do not touch 
the genuine Platonic teaching. 
37 
ParmenideSy Theaetetusy Sophist, Politicus 
connexion with mathematical concepts f'the equals themselves' 74c), 
which is taken up at Parm. 129b, 'the similars themselves'.1 E) The 
Form itself has the character which it implants in particulars. Beauty, 
for instance, is also 'the Beautiful itself, the very perfection of beauty, 
and of Largeness it is said that 'being large, it cannot bear to be small' 
(i02e). F) The extent of the world of Forms is not discussed, but 
mention is made of Forms corresponding to value-concepts such as 
good and beautiful, relations (so that 'Simmias is taller than Socrates' 
is re-formulated as 'Simmias possesses Tallness in relation to the 
Shortness of Socrates', 102b-c), mathematical concepts like quantity, 
length or number (iooe-ioic), and physical substances like snow and 
fire (vol. iv, 357, 359). 
All these aspects of the Phaedo-iheory are discussed in the Par- 
menides, but the omissions are at least equally striking. In the Phaedo 
the doctrine of Forms is unthinkable without the complementary 
doctrines of the human soul (mind) as immortal, periodically  
reincarnated, and an intermediary between the visible world and the 
intelligible Forms to which it is akin G9d). The problem of how our 
minds, tied to bodies and living in the physical world, can have any 
contact with the invisible and changeless, is solved by supposing that 
when out of the body we had complete vision of the Forms, of which 
we may therefore be reminded by their imperfect and impermanent 
embodiments on earth, at first dimly, but by perseverance in the 
philosophic life ever more clearly. Inseparable from all this is the 
constant association of the Forms with value. Not only do the moral 
and aesthetic Forms, as in many other dialogues, have pride of place, 
but the intelligible world is consistently lauded at the expense of the 
physical, and even in the case of Forms which might not seem to have 
any special value, such as equality or size, particulars are represented as 
'wishing' or 'striving' to be as their Forms are, but remaining 'of less 
worth' G4d-75e). 
1 Identified by some with the Form, but cf. Bluck, Phron. 1957, 118, and Cornford, P. and P. 
75: 'quantities defined simply as equal and nothing else'. See also vol. iv, 342-5. 
38 
Parmenides 
Parmenides's questions and objections (ijoa-jSi). With Parmenides in 
the lead, the discussion proceeds on Parmenidean lines, that is, by means 
of dilemmas, demanding a choice between two contradictory theses 
only, with no compromise or qualifications allowed. In the Sophist on 
the other hand it is led by an Eleatic heretic, who is not afraid to criticize 
this procedure of his master on the master's own ground, namely the 
alternatives * being' and * not-being'. This is the first hint that the 
Parmenides is intended as a stimulus to further thought, offering no 
positive result but leading to the Sophist which seeks a solution to the 
difficulties raised by Parmenidean logic. 
(i) Of what things are there Forms? (jjoh-e). When Parmenides has 
verified Socrates's theory as * making a division between Forms on the 
one hand and the things that share in them on the other, so that there 
exists a " Similarity itself" separate from the similarity which we possess, 
and a One and a Many' and so on A30b), his first question concerns its 
scope. Apart from 'what Zeno mentioned', Socrates agrees that there 
are Forms of such attributes as beautiful and good. Concerning natural 
species and substances like man or fire he hesitates, and when it comes 
to 'undignified and worthless' things like hair, clay and dirt he feels 
that to posit Forms of these, besides the visible substances, would be 
absurd. He is, he admits, troubled by the thought that the same rule 
should cover all cases, but retreats to things of whose Forms he feels 
sure, and confines his study to them. Parmenides however attributes 
this to a youthful lack of confidence in his own opinions: as his  
philosophy matures he will cease to despise any of these things. 
In view of the uncertainty surrounding the extent of the world of 
Forms, we may take this as representing Plato's own attitude. Since 
hair and clay are classes of substance with a recognizable form or  
nature,1 it would be only logical to assume a separate Form of each. 
Doubtless there is, but the Forms which interest him as a philosopher 
are not these, but the moral and mathematical, and those of the widest 
1 It may be that, as Crombie says (EPD n, 330), P. regarded clay and hair as 'indeterminate 
objects', 'matter left to its own devices', corresponding to 'no definite and intended character'. 
But (a) he would probably not have ascribed so sophisticated a view to the young S. of this 
dialogue; (i>) that the point lies in worth or dignity is suggested by repetitions of the same 
criticism at Soph. 227b and Pol. 266 d; (c) in any case clay is definable at Tht. 147 c. (On Forms 
of clay and hair see also vol. iv, 549.) 
39 
Parmenides^ Theaetetus^ Sophist^ Politicus 
concepts like Being, Sameness, Difference, Motion, Rest, which he 
deals with in the Sophist.1 The comment allowed to Parmenides hints 
for the first time at something which acquired greater prominence as 
Plato re-thought the theory of Forms in his later dialogues, namely a 
doubt about its universally teleological orientation. Two passages will 
illustrate this: (i) Soph. 227b. Dialectic or philosophical method, in its 
quest for understanding through the detection of affinities, holds all 
pursuits in equal honour. Either generalship or de-lousing may be 
adduced as a species of hunting with no hint that one is more  
contemptible than the other, (ii) Pol. 266 d, after a reference back to the 
Sophist, says: ? philosophical enquiry like this one is not concerned 
with degrees of dignity and does not despise the smaller more than the 
greater, but makes straight for the truth every time in its own 
way.' 
When Parmenides's questions become critical, they do not bear at all 
on the existence of the Forms, but only on their relationship to this 
world and to ourselves (as objects of our knowledge)—a reminder that 
we are listening to a man whose own philosophy allowed the existence 
of the intelligible only, and denied any connexion between it and the 
sensible world. Judge by reason alone: human senses and opinions have 
no validity at all. (Parm. fr. 7, 3-7.) This then is the line of his attack. 
Once again he restates the doctrine, with an actual quotation from 
the Phaedo:2 there are certain Forms, and other things which share in 
them and are called by their names. What shares in Similarity is called 
similar, in Largeness large, in Beauty and Justice beautiful and just. 
Then come the difficulties. 
(ii) What shares in a Form must contain either the whole of it or apart 
Bjoe-j2e). (a) How can it be as a whole in many separate things,3 
when it is a unity? Why not (says Socrates) as a day is in many places 
1 On the extent of the world of Forms, a question which as Goldschmidt truly said ' ne parait 
pas admettre de solution satisfaisante' (Essai 201), some remarks have been made already. See 
vol. iv, 359 (Phaedo), and 548-51 (Rep.), and this vol. 22 (Crat.). For assessments of the 
evidence in both P. and Aristotle see Ross, ? ?? ch. 11, Joseph, K. andG. 65 ff. and the writers to 
whom they refer. In the late Philebus A5 c) P. mentions as examples Man, Horse, Beauty,  
Goodness. From Aristotle we learn that the subject was still unsettled and under lively discussion in the 
Academy. For the pronouncement on it in the Seventh Letter see pp. 407 f. below. 
2 130 c Cf. Phaedo 102 b. 
3 The question whether a Form can either be parcelled out among its instances or exist in each 
as a ^hole is posed again in Philebus A5 b). 
40 
Parmenides 
at once without losing its unity or becoming separated from itself? 
Ignoring this analogy, Parmenides counters with another: you might 
as well spread a sail over a number of people and say that it is one 
whole thing over many. The young and inexperienced Socrates has no 
reply to this, though in fact a material object like a sail is very different 
from a period of time,1 which provides no bad analogy for a  
relationship which, as Plato saw and Aristotle deplored, can only be described 
analogically or metaphorically.2 
(h) Assuming that he has proved his point that a Form must be 
divisible and each particular possess a part, Parmenides goes on to draw 
from it a string of absurdities. Each large thing will be large by having 
a portion of largeness smaller than Largeness itself, ? will be equal toy 
by receiving a portion of Equality less than Equality itself, and finally 
if the explanation of a man being small is that he possesses a portion of 
Smallness, (i) Smallness ('the Small itself K must obviously be larger 
than its part, (ii) an individual is made smaller than before by having 
something added. 
The examples are taken from the Phaedo (i02b-i03a), where it is 
said that if Simmias can be called both big and small, being bigger than 
Socrates but smaller than Phaedo, the explanation is that he possesses 
both Largeness and Smallness, the one in relation to the Smallness in 
Socrates and the other in relation to the Largeness in Phaedo. 
Plato may well have thought that the theory in this form needed  
rethinking, or at least re-formulating, and especially that the notion of 
immanent Forms laid itself open to misinterpretation.4 But Parmenides's 
1 I have no doubt that ????? in ordinary usage meant this, and that references to light or the 
sun are irrelevant. A Greek could of course say ????? ???????? (see LSJ) as we speak of daybreak 
or broad day, but in either language, unless the context demands it, a reader would think of it in 
the sense in which we say that two events happened in different places on the same day. Mrs 
Sprague not unfairly compares Parmenides's argument here with Dionysodorus's analogy  
between Beauty and an ox in Euthyd. (vol. iv, 278). 
2 See further on this Crombie, EPD 11, 330f., 333. 
3 Parmenides uses the abstract noun 'Largeness', but the adjectival forms 'the Equal' and 'the 
Small'. This is not sharp practice on his part, for Plato has always treated the two expressions as 
identical; and it is essential to his theory in the Phaedo that 'the Small' (or the Form Smallness) 
cannot become large in any respect or in relation to anything else. 
4 I cannot here attempt a critique of Fujisawa's thoughtful and challenging article on "?????, 
???????? and idioms of "Paradeigmatism" in Plato's Theory of Forms' (Phron. 1974), in which 
he argues that a distinction between immanent character and separate form, and so between §???? 
and ????????, 'is and will remain ultimate and fundamental in Plato's theory', and that Parm's. 
point depends on confusing them. It would take an article of at least equal length. (But I do not 
41 
Parmenides, Theaetetus, Sophist, Politicus 
criticisms are based on the assumption, drawn from a crudely  
materialistic analogy, that an intelligible Form can be cut up and divided like a 
cake, whereas Plato had written in the Symposium B11b) that a Form 
is eternal and single, and that particulars share in it in such a way that 
while they come into being and perish, it does not increase or decrease 
or change in any way. 
(iii) First regress argument: the largeness of the Large Bj2e-j26). 
Parmenides suspects that the reason for Socrates's belief in a unitary 
Form was this: experiencing a number of things as large, he thought he 
detected a single character in them all, and therefore that 'the Large' 
was a single thing. A glance back at the Meno or Euthyphro shows that 
this is a fair enough summary of the origin of the theory of Forms, but 
what follows? If 'the Large itself is large—and this seems obvious— 
it must belong to the same class as the visible large things, and so it and 
they together share the same characteristic, which on the theory will 
exist separately; but it will also be large, and so ad infinitum. Each Form 
will not be single, but an indefinite plurality. 
This is the argument which has become known as the 'Third Man', 
being one of the arguments described under that title by Aristotle,1 in 
which' man' is substituted for the' Large' of the Parmenides. It involves 
the notion of the self-predication of Forms (as it has been called), and 
(especially since Vlastos's article in 19542) these two, often under their 
abbreviated titles TMA and SP, have become a battlefield for  
commentators. Every possible view has been both asserted and denied by 
scholars modifying not only the views of others but also their own, and 
think he is right in attributing to P. himself the extreme flux-theory of the Tht., p. 53, n. 58.) 
In vol. iv, 353-6, I tried to defend the view that in Phaedo it is the Forms themselves that enter 
into things. 
1 So at least it is generally thought, though Leisegang denied it (RE 2485). For Aristotle's 
evidence see Ross's ed. of the Metaph., vol. 1, 194-6, or Cornford, P. and P. 88-90, and on the 
possible Megarian origin of the argument see Taylor, trans, pp. 21-3, v. Fritz, ?-ESuppl. v, 722, 
Cornford, o.c. 89, Burnet, T. to P. 253 f. 
a Repr. in Allen's SPM. Even then V. could begin by saying that hardly a text in P. had been 
discussed as much in the last forty years as the two passages in Parm. invoking the TMA. He lists 
there 9 * major contributions'. Returning to the subject in 1969 in PQ, he gives 16 (including 4 of 
his own), and this article has already been replied to by S. Panagiotou in PQ 1971. Add Teloh 
and Louzecky in Phron. 1972, and Clegg, Phron. 1973, and a reader will be reasonably well 
equipped to pursue this topic, though he should look also at Crombie's review of Allen's  
anthology in CR 1966, 311 f. The standard works on P. should of course also be consulted. (Vlastos's 
latest and longest list is in his PS, 1973, 361 f. Later come S. Peterson, J ? 1973 and (also relevant) 
D. M. Armstrong, Aust. J. Ph. 1974.) 
42 
Parmenides 
must be borne in mind by anyone reading the necessarily brief account 
offered here. 
That in Plato's eyes Justice itself was just, Piety pious and Beauty 
beautiful, both before he came to separate Forms from their instances 
and in his statements of the theory of separate Forms in the middle 
dialogues, has been noted more than once (see vol. ??, ? i9f., 223,359f.), 
and is inherent in his indifferent use of the substantival and adjectival 
forms.1 This is a survival of the ambiguous Greek use of article with 
adjective whereby 'the hot' appears to refer both to heat and to that 
which is hot, with, in Anaximander at least, a distinct bias towards the 
latter. In any case, whether as paradigm or as 'shared in', it is by 
imparting its own characteristic that a Form is the cause of particulars 
being what they are, and it must, therefore, possess this characteristic 
itself.2 
According to the same middle dialogues, however, it possesses it in 
a peculiar way. In Rep. 5 (vol. iv, 487 f.) Beauty is distinguished from 
its instantiations in this world—sights, sounds and so forth—as being 
the unchanging reality which they fitfully and in a relative sense 
imitate. It is beautiful always, everywhere and absolutely, not beautiful 
in comparison with this but not with that. The Form of Beauty (idea, 
479 a 1) is unambiguously characterized by—is the perfect exemplar of 
—itself. Nevertheless there is this great difference. The ordinary man, 
trusting to the senses, sees only the many beauties of this world, for 
Beauty itself can only be perceived by the mind. So too in the  
Symposium the goal of the philosopher's pilgrimage is itself something  
unchangingly beautiful, in no respect ugly, not to be grasped by the 
senses but only in a flash of mental vision supervening on a strenuous 
course of dialectic. Only that brings knowledge of the truly beautiful.3 
In Phaedo and Phaedrus it is the cause of the beauty of earthly things 
1 For the persistence of this identification of 'universale' with * perfect types' Miss Hicken 
aptly refers to Soph. 256aff., where P. illustrates the point that no Form can stand in a relation of 
sheer identity with its opposite by saying that Movement does not rest. 
2 Ross (PT/ 86 and 88) says that the cure is to realize that the Form is not another thing, but 
an attribute. This would abolish the most distinctive character of the theory of Forms, their 
independent existence, and it is at least doubtful whether P. was ready to do that. That some 
Forms are predicated of themselves is undeniable; e.g. 'the One itself, or Unity, is one. See 
Crombie, EPD n, 347, n. 1, CR 1966, 311. Weingartner makes the point over again, UPD 193. 
3 'True beauty* if you like (and so it is often translated), but only if it is understood that it is 
itself beautiful. 
43 
Parmenides, Theaetetus, Sophist, Politicus 
by being beautiful itself, but its beauty is not of this world nor 
perceptible to the senses, and this is true of Largeness or any other 
Form. 
Largeness indeed is a particularly good example, because, as a purely 
relative term, it could never be mistaken for a sensible attribute. Indeed 
absolute largeness strikes us as an impossibility: one can speak of  
something purely red, but never purely large. We can call a girl beautiful 
meaning only that we admire her looks, with no thought of comparison, 
but if we speak of a large dog or a small elephant we must always have 
in mind a comparison with other members of its class. Comparing the 
dog to the elephant we at once call it small. Yet largeness is a general 
term which can be consistently used in the sense that a speaker and 
hearer understand each other because it conveys the same concept to 
both. In Phaedo-term\no\ogy the largeness in both dog and elephant is 
the same, though the dog possesses smallness in relation to the elephant. 
There is a Form of Largeness in which both share, and though perfect 
justice in this world is, if not attainable, at least not inconceivable, 
unqualified Largeness as a physical attribute is an impossibility. Since 
however it is a legitimate concept marking 'natural divisions', there 
exists a separate Form of it accessible to reason. Obviously, however, 
it is not anything that can be seen, cut up or distributed as the Form 
Whiteness might (erroneously) be thought to be. 
When Plato wrote these dialogues he clearly believed that a Form, 
being incorporeal, was not large in the same sense as a large physical 
object, and its relationship to particulars was not subject to the  
materialistic criticism of Parmenides. This, however, does not settle the 
question whether he has seen logical difficulties in his earlier  
metaphysical doctrine and is expressing his own doubts through the mouth 
of the Eleatic. Is he saying, let us strip off the mystical language of 
Symposium and Phaedrus, the talk of a revelation vouchsafed only to 
initiates, and of the Forms as divine and apprehended by intuition, and 
see what is left if we stick to a logical analysis alone, as the only method 
proper to a philosopher? Certainly the Parmenides breathes an entirely 
different spirit from the central dialogues, but we had better go a little 
further before deciding. 
(iv) Can the Forms be thoughts? {232b-c). Defeated again, Socrates 
44 
Parmenides 
is driven to try the anti-Platonic solution of Antisthenes (vol. in, 214): 
a Form can retain its unity because it is a thought, occurring nowhere 
but in our minds. Parmenides first meets this with an argument from 
his own poem:1 a thought must have an object, and that object must 
exist. When therefore we think of a group of things as having a certain 
common character, there must be not only a universal concept in our 
own minds but a single reality corresponding to it, the character or Form 
(idea, eidos) itself. By this argument Parmenides answers not only 
Antisthenes but also, in advance, the long line of interpreters who have 
supposed the Platonic Forms to be thoughts in the mind of God.2 Even 
God can only think of the Forms because they are there. This is 
unambiguously stated in the Timaeus. So far, the Parmenidean position 
was adopted by Plato himself, as appears from Rep. 476c: 'Does a 
knower know something or nothing?' ' Something.' ' Something that 
exists, or not?' 'Something that exists. How could he know anything 
non-existent?' It is therefore to him a legitimate proof that the Forms 
are not mere concepts, but exist independently of our thought of them. 
His modification of it, as we have seen (vol. iv, 487 ff.), was to allow for 
particulars as a class between existence and non-existence and cognized 
by a faculty (belief or opinion) between knowledge and ignorance. On 
the view taken by Parmenides here, Forms exist, each with the  
properties of his own One (eternal, changeless, single, indivisible, isolated, 
grasped by thought alone), but nothing else exists, and if it did it could 
enter into no relation with such an intelligible unit. 
Parmenides also produces a second objection. If, as Socrates says, 
'the other things' partake of the Forms, either each will be composed of 
thoughts, and everything thinks, or else they are unthinking thoughts. 
One's immediate reaction is to say that it is not a thought (concept) that 
thinks, but the mind which forms it. If I think of something existing 
outside me, there are three factors involved: a thinking mind, the 
concept which it forms, and the reality of which it is the concept. 
Parmenides has used the Greek word noema, in form a passive noun 
from the verb noein (to apprehend by thought), but from Homer  
onwards commonly used in an active sense, to signify an act of thought 
1 Frr. 3 and 8, 34-6 DK. On these see vol. 11, 14, 39-41. 
2 See Audrey Rich, 'The Platonic Ideas as Thoughts of God', Mnem. 1954. 
45 
Parmenides, Theaetetus, Sophist, Politicus 
or even the thinking mind.1 As Grote truly said, the argument is not 
easy to follow. It can hardly be reproduced in English, but to a Greek 
it could seem a fair objection to the view (which Plato had no wish to 
defend) that the Forms are no more than concepts in the mind.2 
(v) Second regress argument: Forms as patterns or paradigms B32 c- 
33a). Baffled again, the youthful Socrates tries the other main  
explanation of the relationship which was accepted without question in the 
Phaedo. The real meaning of'participation' is3 that Forms are a sort of 
patterns fixed in the real world and particulars resemble them and are 
made in their image. Parmenides replies that this relationship is 
reciprocal: in so far as a particular resembles the Form, the Form must 
resemble the particular; but if two things resemble each other they do 
so by sharing the same character, and what is this character but a Form? 
It follows that nothing can resemble a Form, nor a Form resemble 
anything else. Otherwise a second Form immediately appears, and if it 
resembles anything, then another, and this series will be endless. 
This argument depends for its force on the question whether the 
resemblance assumed between particular and Form must be  
symmetrical, as are resemblances between particulars. If ? is like b in being large, 
b must be like a in the same respect. If the explanation of their  
resemblance is that both resemble the Form of Large or Largeness itself, 
does it too resemble them in the same way? There has been much 
dispute over this. Taylor and Cornford (following Proclus) said no. 
The relation of sensible particulars to a Form is that of copies to an 
original, and that is not simply one of likeness. The reflection of a face 
in a mirror is both like the face and a copy of it: the face is like the 
reflection but not a copy of it. Others (e.g. Hardie, Ross, Ryle, Owen, 
1 In Greek his alternatives are ? ... ?? ???????? ??????? ????? ??? ????? ????? ? ??????? 
???? ?????? ?????. Both ????? and ?????? are in form passive, but as commonly used active, 
??????? meaning * unthinking*, not, as its form suggests, 'unthought'. In many places ????? 
could as well be translated 'mind' as * thought*. See Xenoph. fr. 23, Parm. 7.2 and 16.4, Emped. 
105.3, 110.10, Aristoph. Clouds 229. 
2 Of several interpretations Peck's in PR 1962, 174-7, is especially interesting, though in view 
of 132b 4-5 it is difficult to agree with him that S. is not temporarily abandoning the  
transcendence of the Forms. Cf. Johansen, CL et Med. 1957, 7 n. 14. 
3 This disposes of the idea that * participation * and * imitation* might be different relationships, 
upheld by P. at different stages of his thought. Cf. Cherniss in SPM 362-4, and especially Arist. 
Metaph. 991 a 20: 'To say that they are paradigms and that other things share in them is empty 
talk and poetic metaphor.' 
46 
Parmenides 
Runciman) think this reply vitiated by its reliance on the words 
'simply' and 'merely'. Granted that the relationship is not merely one 
of likeness, it still involves likeness. A model and its copies are related 
by resemblance even if that is an incomplete account of their relation.1 
I believe myself that Plato did not admit the objection, and that his 
defence would lie in the non-sensible nature of the Forms. I have 
referred to this already in the context of the somewhat mystical 
language of the Symposium or Phaedrus, but in the Cratylus he has 
given more philosophical expression to this essential difference between 
a Form and its physical manifestations. Runciman has written (in SPM 
i58f.) that the paradigm-theory reduces a Form to the logical status of 
a particular. ' If whiteness is white (which must follow if white objects 
are white by resembling it) then whiteness is one of the class of white 
objects.' Now in the Cratylus Socrates's position is that we must know 
the Forms of things, through which they have their being or essence 
(ousia), before we can communicate by applying names to them (p. 28 
above). At 423 c-e he says that the art of naming does not consist in 
trying to reproduce in words actual sounds, shapes and colours. That 
belongs to music and the graphic arts. But sound, shape and colour 
each have an ousia in contrast to their visible and audible  
manifestations.2 Ontologically at least, the Form is not reduced to the status of a 
particular. It may be, as the same scholar remarks with Aristotelian 
austerity, that nothing could resolve the difficulties raised in the 
Parmenides because 'the theory of forms is logically unsound', but for 
Plato at least, the status of an intelligible could never be on a par with 
that of a sensible.3 
1 Taylor, PMW 358; Cornford, P. and P. 93f.; Hardie, Study 96, Ross, PTI 89; Ryle in 
SPM 105; Owen, ib. 319 f.; Runciman, ib. 158. 
2 It is perhaps useful to remind oneself here of the course of the discussion in Meno. For ousia as 
a transcendent Form see Parm. 133 c. It was one of Aristotle's objections to the theory of Forms 
that it made the substance of things exist apart from as well as within them (Metaph. 991 bi). 
3 Cf. my review of Wedberg's Plato's Philosophy oj Mathematics, Philosophy 1957, 370.1 hope 
I have now answered Weingartner when he writes (JJPD 192): 'The unacceptability of SP is 
even more obvious when we consider such forms as that of Noise (listen to it I) and of Visibility 
(look at it now!).' It should give some support to Peck's view in PR 1962 'that Forms are 
ontologically superior versions of a quality which should be referred to as, for example, the large 
(intelligible), while a particular should be referred to as the large (visible)'. I take this summary 
from Clegg's article in Phron. 1973, 35. His own opinion on p. 37, that * Participation in a 
Form guarantees that what does the participating is without class-membership' because it is  
imperfect, seems topsy-turvy. Class-membership is just what participation in the same Form does 
47 
Parmenides, Theaetetus, Sophist, Politicus 
A few more words are needed on transcendent and immanent Forms. 
When we first met them in the Phaedo (and I intentionally repeat here 
a part of vol. iv, 354f.) I took the view that the Largeness in us was the 
Form itself which also existed beyond, and that there was no need to posit 
a third ontological level between Forms and particulars. This seemed 
confirmed by the fact that it is purely large, never admitting any  
admixture of its contrary as concrete individuals do, and so (one would think) 
in no way imperfect. Ross however supposed the immanent qualities to 
be themselves imperfect copies, and more recently Rist has written that 
the largeness in the particular is 'of an ontologically defective kind'. 
He adds later that' Whiteness is the cause of white in white particulars; 
it is not itself the whiteness in those particulars.'1 Yet the Phaedo says 
it is by its presence in particulars that the Form can act as a cause. 
When writing the Phaedo, as I have said, Plato may not have been 
clear in his own mind about this, but the nearest approximation to his 
thought at that stage seems to be as follows: Whiteness is an intelligible 
(not visible) Form. When it enters a material object (say a face), its 
combination with body produces visible whiteness, an imperfect  
imitation of the transcendent Form in the only medium in which material 
objects can reflect it. The face, which was never perfectly white, may 
turn red by 'receiving' {Phaedo i02d-e) Redness instead of Whiteness, 
but Whiteness, whether 'by itself or in us, will always be Whiteness 
and nothing else. 
' It may be said that this is not 'self-predication': the Form has not 
the quality that it is, for invisible, intelligible whiteness is not white in 
the only acceptable sense of that word, if indeed it means anything at 
all.2 That has at least the advantage that it relieves Plato of the ' Third 
Man' argument.3 Nevertheless, as we have seen time and again, for him 
guarantee. One might almost say that to explain class-membership, to answer the question by 
what right we group certain individuals together in a class and give them the same name, is a 
raison d'etre of the theory of Forms. 
1 Ross, ? ?? 3o, Rist, Philologus 1964, 221 and 223. Likewise Cornford says (P. and P. 78) 
that the tallness in a person 'is not exempt from all change*. This directly contradicts what S. says 
in Phaedo, It is its possessor who is not exempt from change. 
2 Nor is it simply the concept of whiteness, *a thought in the mind*. That interpretation, 
already rejected in this dialogue, is not Plato's. 
3 The main thesis of J. N. Findlay's book Plato, the Written and Unwritten Doctrines is that 
Plato's ontology is not in fact dualistic. There are not two parallel kinds of being: only the Forms 
exist. Consequently, he argues, all arguments of the * Third Man' type must fail. 
48 
Parmenides 
Beauty was the perfection of what is always and in every respect 
beautiful, Largeness the large par excellence and so on; and he is 
beginning to see that such a doctrine has certain logical drawbacks. 
(vi) The Forms unknowable to us and we to GodB 33 b-34 e). Following 
up his rejection of any kind of participation of sensible particulars in a 
Form, Parmenides's next point is that it would be difficult to argue 
against anyone who claimed that Forms, 'being what we say they must 
be', will be unknowable. If they exist' by themselves', i.e. not in our own 
world (a genuine tenet of the theory of Forms to which Socrates  
immediately agrees), they must be related only to each other, not to the copies 
—or whatever we like to call them—in our world. Similarly, things 
in our world which indicate a relationship, though named after the Forms, 
can only be related to each other. Asked to explain further, he offers the 
illustration of a master and his slave. The one is not slave of the Form 
of Master, nor the other master of the Form of Slave.1 Each is a man, 
and his relationship is with a man. But Mastership itself exists in relation 
to Slavery itself, a relationship entirely within the world of Forms. 
Now knowledge (as Aristotle said, Cat. 6b 5) implies a relationship, 
being necessarily knowledge of something. In and by itself, then, 
Knowledge will be of Reality itself, and its branches, the Forms of the 
sciences, will have as their objects the varieties of Reality. Therefore 
if we have no part or lot in the Forms, which are not in our world, and 
every Form is known by the Form of Knowledge, none of the Forms— 
the Beautiful itself, the Good itself and the rest—can be known by us. 
Worse still, no god or gods can have knowledge of us and our world 
nor be our master. Knowledge itself is perfectly accurate, and if any 
being has it, it must be a god, but from what we have agreed about 
Forms having no reference to our world, it follows that Knowledge in 
the gods' world cannot be knowledge of us,2 nor their Mastership 
exercised over us. 
This argument is generally dismissed as fallacious, especially the part 
1 ????? ????????, 6 ???? ????????, but in the next sentence we find the abstract noun, 
???? ? ??????????. P. intended no distinction between these expressions. All occur interchangeably 
elsewhere as synonymous with a Form (?????). 
2 That God, the ultimate cause of everything in the physical world, had no knowledge of that 
world, was the serious view of Aristotle. It would detract from his perfection, and the world was 
sustained in being (not brought into being, for it was eternal) by its own inner drive towards the 
perfection of form represented by God. 
5 
49 
GHG 
Parmenides^ Theaetetus, Sophist, Posticus 
about the gods' knowledge ('unwarranted' Ryle, 'meaningless ... a 
worthless fantasy' Ritter). Cornford said it confused a Form with a 
perfect instance of it.' The form itself... cannot know anything. '* This 
reopens the whole question of 'self-predication'. If Plato said that 
Beauty was perfectly beautiful, he was bound to say that Knowledge 
was knowing, and he could only avoid these errors by the dualistic 
metaphysics of the Phaedo. Such a two-world theory was impossible 
for Parmenides, for whom the only alternatives were' It is' or' It is not', 
and the latter was inconceivable. He is speaking in character. His  
pressure all along has been against any sort of connexion between the real 
world and the sensible, which in his eyes of course was non-existent. 
So now he uses his familiar weapon, the 'either-or' dilemma. Either 
Forms are outside our world and ourselves or they are within: there is 
no middle course. One misses immediately Plato's conception of the 
human soul as the epistemological link between the visible and  
intelligible orders, as 'akin to the Forms'. For Parmenides there are only 
two faculties: logos or nous which grasps the unity of reality, and the 
sense-organs whose fantasy of a world of plurality and variety is 
utterly unreal. Plato's suggestion that the senses might take us the first 
steps on the way to an understanding of the intelligible (Phaedo 74a-b), 
the idea in the Phaedrus that the human mind can grasp the unity in 
the plurality, the universal in the particular, and so begin the process of 
recollection of the Forms—all this is foreign to the elementary logic of 
Parmenides, who is arguing from his own premises.2 
Conclusion on Part One 
Why did Plato write it? Because in the first place, I suggest, his own 
system, with its equation of the real with the intelligible, was firmly 
rooted in the Eleatic's. He had however introduced substantial  
modifications, not glancing back to Parmenides as he did so, but seized by 
the inspiration which fired him to the amazing intellectual and  
imaginative flights of the Phaedo and Phaedrus. Now he feels the necessity to 
pause and take stock, to clarify once for all his position vis-a-vis 
1 Ryle in SPM, 105, Ritcer, Essence 124, Cornford, P. and P. 98 f. For an able defence of the 
argument see Bluck, CQ 1956, 31-3. 
2 A different approach to this argument has been made by J. W. Forrester in Phron. 1974. 
50 
Parmenides 
Parmenides. Parmenides had oversimplified and his conclusions could 
not be the last word. But Plato himself had perhaps ignored this 
simple logic too much, and his own doctrine of Forms, and especially 
the questions of their relation to particulars and of our knowledge of 
them, needed a sober reappraisal and overhaul. 
Being Plato, he puts the critical part of the task in the dramatic form 
of a personal encounter with his great predecessor. Chronology  
demanded the fiction that his mouthpiece Socrates had evolved the full 
Platonic theory as a very young man, but this had the advantage of 
offering Parmenides only the mildest opposition. Before it is  
rethought, the theory of Forms must be submitted to the most rigorous 
examination compatible with the fundamental assumption (which he 
shared with Parmenides) of a stable and intelligible reality. Here he 
points out difficulties. The positive side of the process is left to the later 
dialogues in the group. For instance, in the Sophist B49cff., pp. 142ff. 
below) the soul is restored to its place in the real world, but in terms 
very different from those of the Phaedo. 
On the unknowability argument Parmenides chooses his words  
carefully. He does not say it is irrefutable, but only that to show that it is 
wrong would need a long and abstruse argument with an opponent 
both experienced and gifted: and he concludes by saying that in spite 
of this and many other difficulties, and though it may need a genius to 
maintain the existence of Forms, to deny it would rob thought of all 
direction and make rational discourse impossible.1 The bafflement which 
1 Rist (CQ 1970, 227) says that the only demand here is for Forms as class-concepts or 
universals. 'Philosophy . . . operates with general propositions, and if particulars cannot be 
classed ... (whether or not the classes are Platonic Forms), then thought is at an end.'' There is 
no assertion by Parmenides that philosophy is impossible without separate Platonic Forms, there 
is an assertion that philosophy is impossible without ????.' Weingartner makes a similar point 
{UPD 149) as an argument that P. in Parm. abandons the notion of Forms as paradigms. Corn- 
ford on the other hand (P. and P. ioo) saw Parmenides as accepting the full Platonic view. Since 
Forms are necessary as ' objects on which to fix our thoughts, and as constant meanings of the 
words used in all discourse', they 'must not be wholly immersed in the flow of sensible things. 
Somehow they must have an unchanging and independent existence, however hard it may be to 
conceive their relation to changing individuals' (my italics). 
I believe Cornford is right. That 'an essence all by itself (????? ???? ???* ????? 135 a) should 
be nothing more than a 'common factor' in particulars (Rist 229) is utterly at variance with the 
way the phrase has so far been used in the dialogue. (Cf. esp. 133c 2-6.) I cannot think that P. 
would suddenly have expected his readers to see that Parmenides was now abandoning the sense 
given to ????? in all his previous arguments, which depended for their force on its separate and 
independent existence. 
51 
5-2 
Parmenides, Theaetetus, Sophist, Politicus 
Socrates now feels is simply evidence that he has tried to run before he 
can walk. He cannot expect to seize the truth about Forms like Beautiful, 
Just and Good unless, before he is much older, he submits to a tedious 
training in what is commonly dismissed as useless talk.1 
Transition to Part Two (i35d-3jc) 
What manner of exercise is needed? asks Socrates. The manner 
exemplified by Zeno's arguments which they have just heard, with one 
difference. Socrates himself had suggested confining the discussion to 
Forms, objects of reason, and ignoring the objects of sense ('Yes, 
because I don't see any difficulty in sensible things having contrary 
properties'), and they should continue to do so.2 Also, one must  
consider the consequences not only of any hypothesis being true, but also 
of its being false, e.g. in Zeno's case not only the hypothesis 'if there is 
a plurality', but also 'if there is not a plurality'. One must ask what are 
the consequences in either case for the many, the one, and their mutual 
relationships; and so also with similarity and dissimilarity, motion and 
rest, birth and destruction, and being and not-being themselves. 
' In a word, whenever you suppose that anything exists or does not exist or 
has any other character, you ought to consider the consequences with 
reference to itself and to any one of the other things that you may select, or 
several of them, or all of them together; and again you must study these 
others with reference both to one another and to any one thing you may 
select, whether you have assumed the thing to exist or not to exist, if you 
are really going to make out the truth after a complete course of  
discipline. >3 
Socrates, appalled at the magnitude and difficulty of this programme, 
begs for an illustration of the method at work, and Parmenides is  
prevailed upon to apply it to his own postulate 'about the One itself and 
consider the consequences of the existence or non-existence of its sub- 
1 ?????????, a charge brought against both Socrates and Plato. See vol. iv, 431 n. 3, 499 n. 4. 
P. is probably thinking of Isocrates. Cf. his In soph. 8, Antid. 262. 
2 Taylor, Cornford and Runciman (SPM 161) speak of not confining discussion to visible 
things, but the Greek plainly says that they are to be excluded altogether. This would naturally 
be approved by Parmenides. 
3 i36b-c, trans. Cornford. I should prefer to render the last few words: 'if you are going to 
carry out a complete course of training preparatory to discerning the truth properly'. The aorist 
participle, as often, carries the weight of a main verb. 
52 
Parmenides 
ject. This examination occupies the whole of the rest of the dialogue, 
which from now on changes its character completely. It proceeds by 
question and answer, but the youngest present (Aristoteles) is chosen 
for respondent as 'likely to give least trouble', and the exposition could 
as well have been continuous. The narrative form is quietly dropped. 
With what expectations should we approach this second part? First, 
it is offered simply as one example of a series of dialectical exercises 
which Socrates should undergo while still young A35 d 5-6). As verb 
or noun, the word 'exercise' (?????3?, ????????) is used five times to 
describe it, and it is strange that some have seen in the coming section a 
promise of more. It is to be a training through which Socrates must 
'drag himself A35d3) before he can hope to see the truth.1 Secondly 
it is said to be of the same type as Zeno's. His procedure was to assume 
that only two opposed hypotheses are conceivable, and leaving one 
aside, defend it indirectly by showing that the other led to absurd  
consequences. The flaw in this was that both hypotheses might be  
untenable, being wrongly or incompletely formulated (as indeed were the 
'It is' and 'It is not' of Parmenides; see vol. 11, 73 f.), and as an exercise 
they are now to apply the deductive procedure to both sides of the 
Parmenidean antithesis, the hypothesis of the One as well as its  
contradictory. It can best be described as an exercise in dialectic in the 
Aristotelian sense, useful primarily as mental training, secondly to meet 
opponents on their own ground, and finally for progress in philosophy 
itself because 'the ability to raise difficulties on both sides of a question 
makes it easier to detect truth and error in every case'.2 It is in this sense 
that Plato's Parmenides says exercises like this are necessary if the truth 
is not to escape Socrates. That he should simply raise the aporiai is 
1 It may seem presumptuous thus summarily to take one side in a dispute which was raging 
in the time of Proclus and has on the other side such names as Hegel, K. F. Hermann and Zeller, 
as well as many more recent scholars. (See Friedlander, PL in, 504^, n. 23.) A good defence of 
the view that Part 2 is more than mental gymnastics is Runciman's in SPM 168-71 (against 
Robinson), and it is also Cornford's position in P. and P. What seems to me incontrovertible is 
that 135 C-36C contain a promise of ???????? and nothing more, not for instance (as Brumbaugh 
puts it, P. on One 189) 'an indirect proof that the theory of forms is a necessary presupposition 
of understanding anything at all'. If I understand Zekl's work rightly (his long and complex 
sentences can be hard going for a non-German) this is his conclusion too, that (as he says at the 
end of his introduction, Parm. p. 14) properly analysed and assessed, the dialogue 'becomes 
decidedly what its second part expressly claims: a lesson and an exercise in thinking'. 
2 See Arist. Topics 1 ch. 2. 
53 
Parmenides, Theaetetus, Sophist, Politicus 
appropriate. To tackle them is left to his 'more moderate' follower 
{Soph. 216b) in the Sophist. 
Part Two (i3jc-66c) 
The plan of the exercise is to take the Eleatic hypothesis of Unity and 
follow out the consequences of its being (a) true and (b) false, in each 
case considering the effects on the One itself and 'the Others'.1 Its final 
conclusion is (and these are the last words of the dialogue) : 
Whether One is or is not, it and the others, in relation both to themselves 
and to each other, are and are not, and appear and do not appear, everything 
in every way. 
This sentence is reminiscent of nothing so much as the riddle of the 
eunuch and the bat in the Republic, nor is the resemblance fortuitous. 
The riddle was quoted as an illustration of ambiguity, and of the unreal 
dilemma brought about by asking the incomplete question ' Is it or is it 
not?', without allowing for a middle status between being and not-being 
which is in fact that of the whole sensible world. (See Rep. 479 b-d.) 
By laying the emphasis on the Unity of the One, Parmenides deduces 
that neither it nor the Others can have any properties, be in any state, or 
in any relation to themselves or anything else, or even exist. By starting 
from its existence (which immediately introduces a duality, Unity and 
Existence) he deduces that both it and the Others are an infinite  
multitude, with both of any pair of contradictory attributes, in both of any 
pair of contradictory states (e.g. at rest and in motion), and in  
contradictory relations (same and other, like and unlike, equal and unequal etc.) 
to themselves and anything else. Equally disconcerting conclusions are 
drawn from the hypothesis that 'the One is not' and 'it is not one'.2 
1 Burnet (T. to P. 262) writes that 'the discussion is about forms alone, and we are expressly 
warned against the idea that "the rest" of which he speaks are the things of sense A35 c). They 
are Just the other forms.' For Cornford the terms 'One' and Others' are 'blank cheques' 
(P. and P. 113) until a particular hypothesis makes clear the sense in which they are there being 
used. Thus in the first two hypotheses they are 'sensible appearances', 'physical bodies' (pp. 157, 
203 f.). This variety of views emphasizes the studied vagueness of the language which alone makes 
the contradictory conclusions possible. Similarly some (Ryle, Runciman) have thought that' the 
One' is throughout the Platonic Form of Unity, others that it is not. 
2 A full summary of the 8 (or 9) arguments will be found in Burnet, T. to P. 264-71. Brief 
and clear is Hamlyn in PQ 1955, 298 f. Burnet's section on Parm. makes perhaps the best case 
for regarding part 2 as a polemic against the use of Parmenidean postulates by the Megarians. 
A 'map' of the arguments is also provided by Owen in Ryle, 349-62. 
54 
Parmenides 
' The key to the understanding of the second part must be sought in 
the unmistakable ambiguity of the hypothesis, " If there is a One".' So 
Cornford, and Crombie emphasizes 'the complete vagueness with 
which the topic to be discussed is introduced'. ' The meaning of the 
essential terms shifts as the argument develops.' Without this ambiguity 
and lack of precise definition the arguments could not proceed to their 
mutually contradictory conclusions. We may note, first, that this lack 
of definition, the incompleteness of the predicate in pronouncements 
like 'It is', was a mark of Parmenides himself.1 Secondly, as Plato 
showed in the comedy of Euthydemus, it was adopted by the Sophists 
as the basis of the logical trickery by which they confused their  
opponents and upheld the rhetorical thesis that' on every topic there are two 
arguments contrary to each other'.2 'Both and neither', the triumphant 
cry of Dionysodorus (Euthyd. 300 d), is the conclusion which  
Parmenides is made to reach in this dialogue.3 Gorgias in On the  
Nonexistent showed that by Parmenidean logic one could as easily prove ' It 
is not' as 'It is'.4 The ambiguities were perfectly plain to Plato,5 yet on 
that very ground Cornford denied that he was 'consciously playing on 
these ambiguities to construct a string of sophisms'. 'The student is 
expected to infer' the ambiguity, and on this understanding the  
arguments 'cease to be either fallacious or meaningless', being in fact a 
valid, indeed brilliant, refutation of Eleaticism. As evidence that Plato 
would deem it beneath him to construct sophisms of this sort, Cornford 
quotes the expression of contempt for them in the Sophist B59b-c). 
The fact remains that some of the arguments as presented do play on 
1 Vol. 11, 73 ff., comments on this and on Plato's criticisms and more advanced position. 
3 Vol. in, 5of., 316. That the thesis owed its origin to Parmenides is none the less true because 
he himself would not have approved it. Cornford admits both that Parmenides himself confused 
the two senses of 'If One is' and that the eristic Sophists used the ambiguity 'to entangle  
disputants in contradictions or paradoxical nonsense' (pp. 109, in). 
3 Noted by Grote A1, 290 f.), who adds that if the demonstrations in Part 2 had come down 
under the name of Protagoras, Gorgias or Euclides, critics would probably have called them poor 
productions, worthy of men who made a trade of perverting truth. 
4 For Gorgias's work see vol. m, 192-200. A close parallel occurs in Parm. at 162a. It is noted 
by Cornford (p. 226), who thinks of it as 'answering' one of G.'s arguments, but perhaps it would 
be fairer to say that it makes use of it. Brumbaugh (P. on One 21 f., 22 n. 4) sees a complicated 
relationship, a 'double irony' showing that the joke is on G., not Parmenides. 
5 Though there have been sceptics, e.g. Grote A1, 297) and recently Runciman (in SPM 180): 
' It seems improbable that Plato saw at all clearly where and why the arguments of the exercise are 
fallacious.' 
55 
Parmenides, Theaetetus, Sophist, Politicus 
ambiguities and are therefore fallacious and sophistic; and if Plato was 
aware of this, so, one would assume, were Euthydemus and Dionyso- 
dorus. It seems more likely that the visitor in the Sophist can speak as he 
does just because the exercise in such sophistry provided by the 
Parmenides has already shown up the absurdities to which it leads. The 
one dialogue is certainly a preparation for the other. Readers are 
intended to detect the fallacies, but as a training in how to avoid them, 
and as Parmenides himself put it in introducing his account of men's 
false beliefs, 'that no judgement of mortals may outstrip thee'.1 
Some object that to make Parmenides propound arguments which 
are in any case fallacious but, if valid, would undermine his own  
philosophy, is incompatible with the respect in which Plato held him. I have 
remarked already that this respect was not unqualified; and the point on 
which the Eleatic visitor in the Sophist feels bound to contradict his 
father in philosophy, even at the risk of being thought unfilial, is the 
one which is so conspicuously lacking in the Parmenides, namely that 
'is' and 'is not' are not absolute: 'what is not in some respect is, and 
what is, in a way is not' B41 d). Since all that Parmenides offers is a 
training exercise, one out of several necessary before the positive 
search for truth can begin, one might even conjecture that Plato is 
paying him the compliment of himself seeing through the sophistic 
abuses of his central dictum. The dialogue ends abruptly at the  
conclusion of the exercise, and whatever moral Parmenides might draw from 
it remains unspoken. 
A point remains which has been made by Ross {PTI 100), that to 
treat the second part as 'gymnastics' does not imply that nothing of 
value emerges from its arguments. Ross speaks of'positive ideas which 
will fructify in his later thought'.2 We have noticed, too, in the earlier 
1 Parm. fr. 8.61. I differ from Cornford reluctantly, and would direct a reader to Allan's 
defence of him against Robinson in PQ 1955, 373 f. Important for his denial of sophistry in the 
arguments is the statement on p. no that 'Plato usually indicates clearly enough where he is 
passing from one to another sense or aspect of "the One" or of "the Others". But contrast 
p. 217: the contradictory conclusions of hypotheses 1-4 'can be stated thus only because the 
different meanings of the supposition [that there is a One] have been disguised'. For a full 
critique of C. see Robinson, ? ED 268-74. R. adopts the 'gymnastic' view, as does Ross {PTI 
99-101). Both acknowledge their debt to Grote (PL 11, 293 n. A, which also contains an interesting 
discussion of still earlier views). 
a PTI 100. Perhaps even in later centuries. Cf. Runciman on the mathematical proofs that 
can be discerned at 1433-443 and 1493-c (SPM 165). For Plato one might instance 158c!: 
56 
Parmenides 
dialogues, a puckish habit of interspersing serious Socratic or Platonic 
ideas with otherwise adhominem arguments, though the interlocutor or 
audience is unlikely to appreciate them, and they are not followed up. 
On the negative side,' Parmenides enunciates his contradictory  
demonstrations as real logical problems, which must exercise the sagacity and 
hold back the forward impulse of an eager philosophical aspirant' 
(Grote ii, 301). 
Conclusion 
To understand the purport of the Parmenides is very difficult indeed. 
Every possibility has been put forward and rejected in turn, so that any 
interpretation must be offered with great diffidence. 
My own starts from the conviction that if Plato chose to make 
Parmenides the leading figure in a discussion of the Forms, it was  
because he wanted to clear up the relationship between his own doctrine 
and the Eleatic thesis of One Being. To exalt the intelligible as alone 
fully real was an achievement for which, he believed, philosophy must 
be for ever grateful, but at the same time, stated in Parmenides's terms, 
it would have brought philosophy to a halt. Hence his own efforts to 
provide a bridge between being and not-being, knowledge and 
ignorance. Somewhere the two doctrines had to be brought face to face. 
I have tried to show that this is happening here, and it will continue in 
the Sophist. Direct confrontation with the old man himself leads to an 
impasse, but Plato's debt to Eleatic thought appears when he is replaced 
by a less uncompromising representative of the same tradition.  
Parmenides attributes Socrates's discomfiture to immaturity and lack of 
training in argument, and offers a demonstration. For one thing, 
Socrates was certain that Forms could not admit contrary predicates or 
combine with each other. The demonstration 'proves' that they can do 
both. In this and other ways Parmenides performs the necessary  
preliminary operation of reducing Socrates to perplexity {aporia) as the 
mature Socrates did to people like Meno. And like Meno he is the better 
for it. Only in the Sophist, under more sympathetic Eleatic tutelage, do 
we proceed to build on the ground thus cleared, and learn, for instance, 
the unlimited Many acquire limit through association with the One. This suggests the  
Pythagorean notion which according to Aristotle P. adopted in calling his first principles ' the One and 
the great and small' (or 'indefinite dyad'). See Metaph. 987b!8ff. and other passages cited in 
Ross ch. 12. 
57 
Parmenides, Theaetetus, Sophist, Politicus 
that some Forms can combine and others not. A short paragraph from 
the Sophist will illustrate the point that in the Parmenides Plato states 
dilemmas resulting from the original Eleatic thesis and in the Sophist 
suggests solutions on his own lines. 
We must admit that motion is the same and not the same, and we must not be 
disturbed thereby; for when we say it is the same and not the same we do not 
use the word in the same sense. When we call it the same, we do so because it 
partakes of the Same in relation to itself, and when we call it not the Same we 
do so on account of its participation in the Other, by which it is separated 
from the Same and becomes not that but other so that it is correctly spoken 
of in turn as not the Same.1 
The Parmenidean confusion between identity and attribution is cleared 
away in terms of the doctrine of Forms, and by the realization that a 
word may be used in more than one sense. 
On this interpretation the Parmenides is an aporetic dialogue with a 
difference. The early dialogues showed Socrates skilfully reducing a 
respondent (and as he would say, himself as well) to aporia, thereby 
exposing the confusions of thought underlying the popular use of 
language. In the meantime he has become a teacher with elaborate 
positive doctrines about Forms, soul, the physical world and their 
mutual relations. With astonishing artistry as well as flexibility of mind 
Plato now transforms him again, this time into a young man, keenly 
intelligent and eager for truth yet in argument no match for a great 
philosopher, in order to subject these positive doctrines to an  
examination from the other's point of view. 
Prima facie at least, the first part makes some telling criticisms of the 
doctrines in question, and they are never answered. In face of this, some 
commentators have argued that they are not in fact serious, others that 
they are fatal to the Phaedo doctrine and Plato must have known it (or 
alternatively that he failed to realize how damaging they were), others 
1 Soph. 256a-b; see p. 152 below). The translation is M. G. Walker's (PR 1938, 513; I 
have supplied capital letters for Forms), whose thesis is that P. arrives at his solutions in Parm., 
and Soph, is only conveying the same lesson. She quotes Morris Cohen to the effect that P. avoids 
' the indecent confusion at which we arrive if we violate the principle of contradiction and try to 
wipe out the distinctions of the understanding'. I should have said that he intentionally does not 
avoid it in Parm., but does in Soph., and I claim no originality for this. Cf. Brochard, ?ts. de 
Phil. Anc. et Mod. 167: 'Le Parminide pose le probleme dans toute sa difficult^, le Sophiste et le 
Politique en donnent la solution.' 
58 
Parmenides 
again that they did not touch the essence of the doctrine but called for a 
modification which Plato later effected. Most who take the last view see 
the change as a renunciation of the idea of the Forms as transcendent 
paradigms in favour of regarding them as no more than universals, 
stable general concepts.1 Ackrill, upholding this view, says honestly 
that it would be more natural to call it jettisoning the theory than 
revising it. The remark at 135b, he says, 
strongly suggests that what he is now sure of is not that there must be Forms 
as conceived in the middle dialogues, Forms as ethical ideals and as the  
metaphysical objects of intuitive and perhaps mystical insight; what he is now 
sure of is that there must be fixed things to guarantee the meaningfulness of 
talk, fixed concepts—the meanings of general words.2 
That Plato, as a result of his own criticisms here, gave up the doctrine 
of transcendent Forms, is disproved by many references to it, in 
dialogues universally agreed to be later than the Parmenides, which  
contrast, in the terminology of the Phaedo, a world of realities—eternal, 
unchanging, perfect, bodiless—with the visible world of change and 
becoming. It is true that a list of references only (like Runciman's in 
SPM 152) needs careful checking, for a die-hard believer in Forms as 
concepts or common properties might interpret the language of some 
of them in that sense. At Laws 965 b-e, for instance, Saunders in the 
'Penguin' translation gets on well enough with a vocabulary of  
'concept', 'notion', 'common element' (p. 379 n. 3 below). But one can add 
859c, where the language of 'association' and 'sharing' is more 
strongly reminiscent of the Phaedo. The Philebus has several decisive 
passages, as have Sophist and Statesman, and of course Timaeus (if one 
accepts the traditional dating). It is also explicit in the Seventh Letter.3 
1 An early and formidable champion of a change of doctrine after Parm. was Henry Jackson 
in his series of articles iny. of PhiloL on 'Plato's Later Theory of Ideas'. His conception of the 
nature of the change, however, was different, and based on an interpretation of the Phil, which 
has not found general favour. For a criticism see Ross, PTI 133 k 
2 Ackrill in SPM, 206 (my italics). For Rist's view see p. 51 ?. ? above. Abandonment of 
paradigmatism is also argued by Weingartner {UPD ch. 3), and denied by Cherniss {SPM 361 f.). 
Ross {PTI 86) thought P.'s doubt concerned the 'Largeness is large' form of expression. That 
P. did not realize the damaging effect of the criticisms is the view oiRunciman {SPM 151-3). 
Those who think that he neither regarded nor should have regarded them as serious include 
Taylor {PMW^o), Grube (??36), Cornford {P. and P. 95), Field {Phil, of P. 1 iof.), Crombie 
{EPD 11, 332ff.). 
3 Laws 859 ? ??????? ?? ??? ??????? ??????? ???? ???????? ??? ??? ????? ??????? ????. 
(For the bearing of the Laws on this question see also Runciman, PLE 54f.) See also Phil. 
59 
Parmenides, Theaetetus, Sophist, Politicus 
Two passages are especially interesting as explicitly meeting objections 
raised in the Parmenides. One has been mentioned already. Soph. 248 c- 
49d, in coming to terms with the 'friends of Forms', restores soul to its 
place among the real and explains thereby the possibility of our  
knowledge of a changeless reality. Phil. 15a-b takes up the question whether 
a unitary and eternal Form can be scattered among an infinite number 
of generated individuals, or alternatively be somehow separated as a 
whole from itself. It may be added that Aristotle, in his various accounts 
and criticisms of the doctrine, never suggests that Plato altered it in this 
way. Had he done so, the sting would have been removed from most of 
Aristotle's attacks.1 
Having noted all this we may justifiably remind ourselves how much 
of his own doctrine—call it metaphysical, religious, mystical or what 
you like, but at any rate genuine Platonic doctrine—Plato has omitted 
for the purpose of the experiment with Parmenides (pp. 38,43, 50 
above), and we need feel no compulsion to suppose that he has  
abandoned it.2 Some changes might be needed (e.g. in the language of'self- 
predication' or the status of a Form when it has 'entered into' a 
particular), but the cornerstone of the whole, the transcendent, eternal, 
ideal character of the Forms, remained in place. The challenge of 
Parmenides was how to reconcile this transcendence with a form of 
'association' (????????) both with the sensible world (said in the 
Parmenides to be a prerequisite of knowledge) and with each other 
(declared at Soph. 2596-60a to be essential if discussion is to be carried 
on at all). The casual allusion to both in the Republic^ compared with 
their serious examination in these dialogues, shows how far Plato has 
come from the easy, dogmatic assurance of his golden period. The old 
i5a-b, 58a, 59a-c, 6id-e, 62a; Pol. 269c!, 285e-86a; Soph. 2486-490!, 254a; Ep. 7, 3423-d. 
Kucharski's article 'La "th?orie des Id?es" selon le "Ph?don" se maintient-elle dans les derniers 
dialogues?', in Rev. Philos. 1969, is mainly concerned with Philebus. 
1 Chung-Hwan Chen, so far as I know, is alone in doubting that Aristotle attributed ???????? 
to Plato; and he seems to have misunderstood the attitude adopted in Soph, to the 'friends of 
Forms'. (See CQ 1944, 101 with n. 3.) It may be helpful to compare vol. iv, ii7f., 118 n. 1, and 
p. 47 n. 2 above. Ross notes (PT199) that Parm. is the one important dialogue to which Aristotle 
never refers. 
* Some have supposed that the dialectic of the later dialogues replaced the belief in knowledge 
as recollection. But see Gulley in CQ 1954 (esp. pp. 209 ff.) and Rees, Proc. Ar. Soc, suppl. vol. 
1963, 172??. (against Strang). 
3 476 a. See vol. iv, 498. Similarly at Phaedo i02d it is clearly stated that a Form must be both 
transcendent and immanent, with no suggestion that this involves any difficulties. 
60 
Parmenides 
Greek problem of the One and the Many—and we must never forget 
that Plato was in this tradition—was not to be so easily conquered. 
'How', as the Orphic Creator asked, 'shall I have all things united yet 
each one separate?'1 
B) THEAETETUS2 
Connected with this is the hardest and most urgent of all problems, to 
which the argument has now brought us. If nothing exists except 
individuals, and there is an infinite number of them, how can one 
attain knowledge of the infinite? We know things in so far as they are 
one and the same and possess some universal attribute. 
Aristotle, Metaph. 999324-9 
Date. The introduction tells of Theaetetus being carried home dying 
of wounds and dysentery after a battle at Corinth. Two such battles 
come into question, one about 394 or not much later, the other in 369. 
Campbell (Tht. lxif.) argued for the earlier, but the later is generally 
favoured today and much the more probable. The Theaetetus is a tribute 
to his memory, and probably written not long after his death, i.e. 
shortly before Plato's second visit to Sicily. The majority would now 
agree on 369/7.3 In spite of its close connexion with the Sophist and 
Politicus (p. 33 above), some are still so impressed with the novelty 
of the method of collection and division in the Phaedrus that they 
regard its absence from the Theaetetus as sufficient evidence of earlier 
composition. I have already given my opinion that the novelty of the 
method has been exaggerated (see vol. iv, 430 f. and p. 28 ?. ? above), 
and can only record a personal impression that the Phaedrus is lit by 
the same glow as Phaedo and Symposium, a glow which has faded by the 
time of Parmenides and Theaetetus and is not recaptured even in the 
Timaeus. Unless the Theaetetus, as a Socratic and aporetic dialogue, is 
to be put in the early group (and few would wish to do that today), 
I would say, on partly subjective grounds certainly, that all four 
1 Kern, O.F. 165. Cf. vol. 1, 132. 
a For a full discussion of philosophical questions raised by Tht, a reader may be referred to 
McDowell's edition. 
3 E.g. Taylor, PMlV^io, Field, P. andC.'s 70, Jowett's edd. in, 392 n. The case for the later 
battle was first argued by Eva Sachs, De Th. A914), 22-40. (Cf. vol. in, 499 and vol. iv, 52.) 
For earlier disputes about the date see her notes to pp. 18 and 19. Dies remained agnostic (Autour 
de P. 247). 
6l 
Parmenides, Theaetetus, Sophist, Politicus 
dialogues in this chapter were written after the great middle dialogues 
which include the Phaedrus.1 
ADDITIONAL NOTE ON DATING THE THEAETETUS 
Some would explain the Socratic character of the dialogue by the theory that 
most of it was written comparatively early and what we have is a revision by 
Plato of an earlier edition. It is largely a matter of internal indications and 
personal impressions, and was conjectured even before the discovery of part 
of a papyrus commentary published by Diels and Schubart in 1905 which has 
been thought to furnish some external confirmation. This rests on the fact that 
the commentator mentions the existence of another, * rather frigid' proem 
beginning * Boy, are you bringing the dialogue about Theaetetus?' If this 
opening also is by Plato (and who would forge it?), the most obvious occasion 
for replacing it by the proem in our manuscripts would be the death of 
Theaetetus, from which it is concluded that the main dialogue, or much of it, 
was written before he died. (See Cornford, PTK 15.) The best defence of 
this theory is by Popper (OS 321 f.), who sees signs in the dialogue itself that 
it was written earlier than the Republic. It may be correct, and should 
certainly not be passed over even if I am not personally convinced. Popper 
adduces a number of arguments in its favour, of which I will only point to 
two that seem to me dubious. 
A) He takes two passages of Aristotle, which ascribe to Socrates the 
invention of induction, and mention his profession of ignorance, to be 
allusions to the Theaetetus. But both these historical facts may be illustrated 
from other dialogues. The profession of ignorance suggests most strongly 
the Apology, but also Symp. 216 d, Charm. 165 b and other places. 
B) In the proem Euclides says that Socrates repeated to him the  
conversation with Theaetetus, that as soon as he got home he made notes of it, and 
that on subsequent visits to Athens he verified some points with Socrates 
himself. Popper claims that this contradicts the statement at the end of the 
dialogue that Socrates's trial was already imminent, which would leave no 
time for such visits, and suggests that it is a relic of the earlier version  
overlooked or ignored by Plato in his revision. As to that, however, see p. 64 
?. ? below (written before I looked at Popper's arguments). 
1 Of recent writers, Robinson (Essays 58) and De Vries (Phdr. 11) agree with Von Arnim 
that Tht. is earlier than Phdr. Stylometry may be a fickle guide, for reasons given by Cornford, 
PTK 1. It led Campbell (Tht. lv) to put Tht. 'between the Phaedrus and Republic*, a result which 
modern admirers of his pioneer work in this field seem content to ignore. On some points the 
'infinitae disceptationes' which Apelt noted in 1897 are still with us; e.g. on whether Tht. was 
completed long before Soph, was composed, contrast Cornford (I.e.) and Ritter (Essence 28). 
That Tht. itself was composed over a considerable period is of course possible. 
62 
Theaetetus 
Dramatic date. At the very end of the main dialogue Socrates casually 
mentions that he must leave for the King's Stoa in connexion with the 
indictment of Meletus. The date is therefore 399, and his trial and death 
are near. (Cf. vol. iv, 102.) His hearers would soon see in reality what 
is described in the dialogue (i73c-75b), the relation of the philosopher 
to the practical world and his behaviour in a court of law. 
Characters. Euclides and Terpsion from Megara were intimate friends 
of Socrates, present at his death (Phaedo 59c). Of Terpsion nothing 
more is known. For Euclides and his philosophy see vol. in, 499-507. 
That such an intimate friend of Socrates should be keenly interested in 
one of his conversations as recorded by another is natural, and need not 
mean that the main dialogue contained reflections on Megarian doctrine, 
but for internal evidence see Campbell's edition, xxxv-xxxviii. 
Theaetetus of Athens, a friend of Plato, became one of the most 
brilliant mathematicians of his generation. Only a boy at the time of the 
dialogue, he receives unstinted praise for his intellectual curiosity and 
promise from both Socrates and his master Theodorus. His death of 
wounds and illness must have struck him at the age of 48-50. Of the 
older mathematician Theodorus of Cyrene, who taught both Plato and 
Theaetetus, the dialogue itself tells us much: his work on square roots, 
his early abandonment of general philosophy to concentrate on  
geometry, his friendship with Protagoras.1 The presence of the Younger 
Socrates is also mentioned, though he remains silent—an additional 
indication that Theaetetus, Sophist and Statesman are to be read as a 
continuous series, for his presence is again mentioned at Soph. 218 b, 
and in the Statesman he replaces Theaetetus as chief respondent. He too 
is a historical figure, criticized by Aristotle for treating physical beings 
like mathematical abstractions,2 and pretty certainly the Socrates  
mentioned in the Eleventh Letter C58 c!) as prevented from travelling by ill 
1 For Theaetetus see Sachs's dissertation already mentioned, von Fritz in RE 2. Reihe, x. 
Halbb., 1351-72, or more briefly Taylor, PMW 322; and for his mathematics M. Brown in 
J HP 1969, 362 f. Brown refers to earlier studies and gives the evidence for T. having written 
most of Euclid Bk 10. His connexion with the construction of the regular solids is mentioned in 
vol. 1, 268f. For Theodorus, von Fritz, ib. 1811-25. 
2 Metaph. 1036b25 if. 'He thought that man could exist without his parts as the circle without 
the bronze.' It sounds as if Y.S. was using a mathematical analogy in support of the full Platonic 
theory of transcendent Forms. For further details about him see Skemp, P.*s Statesman 25 f. 
63 
Parmenides, Theaetetus, Sophist, Politicus 
health. Socrates makes much of the fact that both the boys have  
something in common with himself, one being his namesake and the other 
resembling him in appearance, but if this has symbolic significance it is 
hard to discover. 
Prefatory conversation. This takes place in Megara, where Euclides tells 
Terpsion how he met Theaetetus, barely alive, being carried home from 
the Corinthian battlefield to Athens. They grieve at the approaching 
death of a man so talented and in every way admirable, and Euclides  
recalls how Socrates had prophesied a brilliant future for him when, 
shortly before his own death, he met and talked with him—still a mere 
boy—at Athens. Terpsion would like to hear what they talked about, 
and this is still possible, for as soon as Euclides got home after hearing 
it all from Socrates he made notes which he afterwards wrote up at 
leisure, checking the details with Socrates on further visits to Athens.1 
They therefore settle themselves, and a slave reads the manuscript. This 
is the only dialogue which is represented as actually read, though in the 
introductions to Phaedo, Symposium and Parmenides Plato has been at 
some pains to authenticate the record, at least dramatically. Also of 
interest is Euclides's remark that he has written it not in narrative form 
as Socrates told it, but leaving out the connecting 'and I said', 'he 
agreed' and so on as tiresome, and casting it into direct dialogue form. 
This is a form which, as we know, Plato sometimes used in earlier 
dialogues, but from now on he gives up the narrative form altogether. 
The Parmenides showed a transitional stage, in which the narrative 
form is tacitly dropped half way through, and it is a fair inference that, 
as has been assumed on other grounds, it slightly preceded the 
Theaetetus. 
Introduction to main dialogue. The Theaetetus is a brilliant adaptation of 
the manner and plan of the earlier dialogues to the more critical and 
probing approach to knowledge of Plato's late maturity. The  
restoration of Socrates to his earlier role, with much of his original personality, 
1 S.'s trial was already imminent, but the Phaedo E9d) tells how his friends used to visit him 
in prison during the month which intervened between trial and execution. As E. was in Athens 
for the final farewell, he would certainly have made several previous visits. For the practice 
among S.'s admirers of recording his conversations see vol. m, 343 f. 
64 
Theaetetus 
shows Plato still anxious to be regarded as the true heir and continuator 
of Socratic teaching.1 In this respect it is a complete contrast to the 
Sophist. Socrates is not just a thinking-machine like the Eleatic visitor, 
but Plato has brought out his character by a number of dramatic 
touches, e.g. the Socratic humour of the midwife analogy, the  
seriousness with which his confession of ignorance is followed up in its  
consequences and the positive value of teaching from that position explained. 
This accords with the philosophical purpose of the two dialogues, the 
one aporetic, setting forth problems, the other didactic, solving them. 
Reminders of the earlier dialogues are many. Socrates is still seeking 
out the most promising of the young A43 d; cf. Charm. 153 d), and is 
introduced to one whose name he does not know (i44d; Lysis 204c). 
The aim is to define a given concept, the respondent at first offers 
instances instead {Laches, Hipp. Maj., Meno, Rep. 1), after which 
several suggestions are considered and rejected and the dialogue 
ostensibly ends in failure. The difference lies in the choice of subject. In 
the previous dialogues certain moral or aesthetic concepts have been 
examined—Goodness, Self-control, Beauty, Justice. As to knowledge 
itself, the current puzzle of whether one can learn either what one does 
not know or what one knows has been made fun of in the Euthydemus 
and answered in the Meno by reference to reincarnation and  
recollection. The Charmides even raised the question whether there can be 
knowledge of knowledge (vol. iv, 160f., 169 f.). In Meno and Republic 
the distinction between knowledge and true belief is drawn and is seen 
in the latter to depend on the supposition of the changeless Forms: 
knowledge is the philosopher's recovery of the eternal realities with 
which we all had direct acquaintance before birth, and the existence of 
which is simply assumed. Now for the first time Plato has chosen to 
make knowledge itself the main subject of enquiry, setting aside for the 
purpose all preconceived ideas such as appear unchallenged in the 
Phaedo-Republic group. Nevertheless he still has his own standpoint, 
and it cannot but show itself occasionally. At one point he even turns 
aside, in what is formally a pure digression introduced on the flimsiest 
pretext, to remind his readers that neither the attack on worldly success 
1 This is perhaps also the purpose of emphasizing, in the preface, the pains taken to ensure the 
accuracy of the report. Cf. Stoelzel, Erkenntnisprobl. 6-8. 
6 
65 
GHG 
Parmenides, Theaetetus, Sophist, Politicus 
in the Gorgias, nor the eschatological beliefs of the Phaedo, nor the 
divine Forms of that dialogue and the Republic, are to be regarded as 
superseded. Yet as we saw from the Parmenides, new problems have 
arisen for him, and in his search for knowledge and its objects he shows 
far more interest than previously in the individuals of the phenomenal 
world. The enigma of the Theaetetus may be illustrated by two  
quotations. To Stoelzel it seemed a work that might have been written for 
his own time A908) 'as a weapon in its fight against materialism, 
sensualism, empiricism and positivism'. Richard Robinson on the other 
hand cites its 'empiricist and subjectivist tone' as something 'definitely 
unfavourable to the theory of Forms'. Against this one may note that 
all the empirical and subjective theories discussed are shown to fail, and 
the dialogue could be regarded as a demonstration of their inadequacy.1 
The question: What is Knowledge?2 'This is just the question that 
baffles me: I cannot sufficiently grasp in my mind what knowledge is... 
Speak out like a man: what do you think it is?' A45c, 146c). Here 
Socrates lays down the topic of the whole discussion. But what are the 
1 See Stoelzel, Erkenntnisprobl. v, Robinson, Essays 42. Although P.'s attitude to the Forms 
when he wrote Tht. is much debated, so far as I know no one believes that he had abandoned 
them. Cooper in Phron. 1970 is emphatic that they are not in Tht., but is not there concerned with 
the wider question. In the past Campbell wrote {Tht. liii) that 'Plato's ideal theory, so far as it is 
allowed to appear in the Theaetetus, deals not with hypostatized entities, but rather with necessary 
forms of thought, which are as inseparable from perception as from reasoning.' But he excepted 
the digression, and even Robinson, who so vigorously opposed Cornford's thesis in PTK that 
the Forms are deliberately held back to show that knowledge was impossible without them, 
grants that there may be allusions to them. (See his Essays, 48. McDowell similarly sits on the 
fence, p. 174.) Miss Hicken in SPM argues that Plato is genuinely baffled, convinced of the 
necessity of Forms yet no longer able to distinguish knowledge from belief by their aid. (Cf. 
Raeder, PPhE, 1905, 283: 'Platon versteht nicht mehr das Band zwischen Idee und Wirklichkeit 
zu knupfen.') Most however would agree with the view put forward long ago by Schmidt in his 
commentary of 1880: 'Since neither sense-perception nor true belief nor finally determination of 
the concept (Begriffsbestimmung) prove to be adequate definitions [of knowledge], nothing else 
can be in Plato's mind, as alone in conformity with his philosophy, but a definition directed to the 
Idea, i.e. the reality of the concept or the real and true Being on which it is founded.' Among 
more recent scholars one may cite Ross (PTI 101, 103), Fowler (Loeb ed. 4), Solmsen (P.*s Th. 
76), Hackforth (CQ 1957, 53 ??., a reply to Robinson), Grube (P.*s Th. 36-8), Cherniss (SPM 7), 
Llanos (Vie], Sof. 35), Runciman (PLE 28f.), Sprute (Phron. 1968, esp. p. 67). For references 
to Platonic Forms in dialogues believed to be later than Tht., see above, p. 59 with n. 3. 
2 An observation of Th. Ebert is worth quoting (Meinung und Wissen 9, ?. 15). Scholars 
speak of P.'s 'theory of knowledge' (or epistemology, Erkenntnistheorie), but 'the inappro- 
priateness of the title lies in this, that with it the genetic interest of modern philosophy in the 
problem of knowledge—that is, the question of the sources of our knowledge . . . replaces the 
question of what it is' (in German its ' IVesen'). 
66 
Theaetetus 
criteria that an answer must fulfil? We are up against Meno's pertinent 
question: how do we know what we are looking for before we know 
what it is? Unfortunately it is not the way of Plato's Socrates to lay 
down criteria before beginning the discussion—they are treated as self- 
evident—but just as the Meno's enquiry into arete turned out to be 
based on the hypothesis that whatever arete was, it must be something 
unfailingly good and beneficial in its effects (87d-e), so certain criteria, 
by which candidates for the name of knowledge are being judged, may 
be gleaned as they are casually dropped in the course of the discussion.1 
Without criteria the suggestions could not be tested and rejected. Thus 
we learn that knowledge must be true and infallible A52 c, i6od, 200 e, 
207c-2oc>b) and its object existing A52 c, 186 c) and stable (there 
cannot be knowledge of the ever-changing, e.g. i82e). It must be the 
result of first-hand experience not hearsay B01 b-c), and it must include 
(though the dialogue ends with the admission that these are not 
sufficient to constitute knowledge) true belief (or recognition) plus the 
ability to give an account {logos) of what one believes or recognizes. 
What has no logos cannot be known B02 d, 205 c). 
For comparison, one may quote what has been called ' the classical 
definition of knowledge' in modern times. Though expressed in various 
terms, it amounts to this: A man knows that/? (p being any proposition) 
if (a) he believes/?, (b) he has adequate evidence for/?, (c)p is true. Thus 
'according to the classical definition, knowledge is justified true belief, 
or true opinion combined with reason'.2 This is closely similar to the 
third of the three definitions which Plato here discusses and ultimately 
rejects, but there is a difference in that the modern definition speaks 
only of knowledge in propositional form (knowledge of facts) whereas 
in Plato it is more like knowledge of things, not 'knowledge that' but 
knowledge with a direct substantival object—knowing a syllable, the 
notes of a scale, a waggon, Theaetetus.3 In fact three kinds of knowledge 
1 Late in the dialogue, at 196c!, S. asks permission to do something 'shameless', i.e. disobey 
his own rule in Meno G1 b) and claim to state a property of something whose definition is as yet 
unknown. 
2 Hilpinen, Synthese 1970, 109^, q.v. for refT. to various twentieth-century formulations. 
3 Cf. vol. iv, 493. It is well known as the difference between Fr. 'savoir' and 'connaitre', 
Germ, 'wissen' and 'kennen*. Once English too could mark in words the difference between 
'D'ye ken John Peel' and 'He wist not that it was true' (Acts 12:9). Some have thought that P. 
marked it by his use of ???????, ?????????? and ??????????, but this is not so; e.g. in the short 
67 
6-2 
Parmenides, Theaetetus^ Sophist^ Politicus 
are commonly acknowledged today, the two just mentioned and  
'knowing how',1 as in knowing a game or knowing one's craft, which involves 
a large element of acquired dexterity, skill or technique, or in morals, 
knowledge how to behave. Such knowledge however is never entirely 
divorced from the other two kinds.2 
These distinctions have not fully come to the surface in Plato, who 
throughout the Theaetetus tends to speak of knowing in terms of a 
verb followed by a direct object—a concrete individual thing or person 
—rather than by the equivalent of'how to* or a proposition expressing 
a fact.3 For this there were more reasons than one. First, he was the 
heir of Socrates, the kernel of whose teaching was that the knowledge 
on which all human excellence depended was knowledge of what  
something was. To 'know justice', in the sense of being able to define it, 
was the only guarantee of leading a consistently just life. (Even now 
Plato preserves the substantival expression so unnatural to us: a 
definition of knowledge is desirable because ' the knowledgeable are 
knowledgeable by knowledge'.4) Here (as Aristotle saw) lay the 
originality of his message, for from Homer onwards Greeks had used 
the words in question (???????????, ????????) to denote practical 
abilities or skills, even bodily skills, rather than intellectual  
understanding.5 For Socrates, as for them, knowledge was the basis of both 
technical skill and general excellence, but whereas others had thought 
of this techne and arete as simply knowing how to act, he believed that 
space between i92d and 193 a he has used all three for the same sort of knowledge, namely direct 
acquaintance. The point has been noticed in connexion with Charm, in vol. iv, 169 n. 1. Cf. 
Runciman, PLE 34f.; Sprute, Phron. 1968, 58-60. 
1 The terminology is Ryle's in The Concept of Mind. 
* Cf. Runciman, PLE 11 f. 
3 A rare exception is ??? ????? at 186 b. 
4 ????? ????? ol ????? A45 d: S. has already equated ???(? with ????????, on which see 
vol. iv, 265). For this form of expression cf. ib. n8f. (Euthyphro and Phaedo) and 189 {H. Maj.). 
5 See vol. in, 450 n. 2, and foil. pp. Examples are collected by John Gould, P.'s Ethics 7ff. 
He errs however in supposing that because this use existed earlier it is also the basis of the Socratic 
conception of knowledge. The end (right action) is the same, but the knowledge leading to that 
end has become something different. G. points out also that Hdt. uses ?????????? of being 
convinced of something which is untrue (p. 10), but this only shows up another difference. For 
S. and P. the object of knowledge must be ?? or ??????. (Cf. Gorg. 454d.) More to the point are 
expressions like ????* ????? in Simonides (PI. Prot. 346 c) and the Homeric ????? ????? (//. 
24.41) etc. but these, as Dodds says (G. and I. 17), illustrate the Greek intellectualist approach to 
an explanation of behaviour (making the Socratic doctrine less paradoxical) rather than a 
* behaviourist' explanation of knowledge. For criticism of Gould see Vlastos, PS 205 ff.; Kuhn, 
Gnomon 1956, 339f. 
68 
Theaetetus 
one could only know how to act if one first understood the nature and 
function of the thing to be made—shoe or shuttle—or the moral ideal 
(e.g. justice) to be pursued. Plato started at least from this Socratic idea 
that knowledge is of' things' (universals), whether justice, courage or 
clay (mentioned exempli gratia at i47a-c), and that what one knows 
one can define by stating its eidos (specific character) and so placing all 
instances of it in their proper eidos (class).1 
Other temptations to extend the model of perceptual knowledge or 
direct acquaintance beyond its proper frontiers arose from Plato's 
native language. One has been referred to already (vol. iv, 493 n. 1), 
namely the interchangeability of the Greek words which without  
context we translate 'true' (alethes) and 'being' (on), though the first may 
qualify a thing, a being, a substance or the like as real or genuine and 
the second a statement as true.2 We may occasionally, and rhetorically, 
speak of'a true Englishman', but we would not say of imitation mink 
that it is not true, or of a liar that he says what is not. Even the Sophist's 
clarification of'being' and 'not-being' did not remove this particular 
temptation. Another lies in the fact that ? know what ? is' could be 
idiomatically expressed in Greek (as Plato often expresses it) in the 
form ? know ? what it is'.3 In the course of the Theaetetus itself 
difficulties (doubtless real to Plato) begin to come to light in the 
conception of knowledge as of things rather than facts. 
ADDITIONAL NOTE: EXEMPLIFICATION 
AND DEFINITION 
It is now commonly believed that Socrates was mistaken in claiming that 
one cannot know what ? is without being able to produce a definition of it: 
that even if a definition could be produced it could not be a means of knowing 
what ? is since, to take the example of knowledge, * one would need to have a 
complete grasp of all the cases of knowledge, and of their relations to each 
other and to everything that is not knowledge, before one could know that 
the definition was correct'. (Quotations are from Bambrough, Reason, 
Truth and God ?^?.) Wittgenstein is thought to have delivered a fatal blow at 
Socrates's demand for definitions in The Blue Book (p. 20), by turning the 
1 See further on this pp. ii2f. below. 
2 E.g. ??? ????? ????? ??????, Hdt. 1.95*116; ?? ???? ???????????, Thuc. 7-8.2. 
3 See also McDowell's remarks on pp. 188 and 192^ of his Tht. 
69 
ParmenideS) Theaetetus, Sophist, Politicus 
tables on him. Socrates will not accept, even as a preliminary answer to the 
question 'What is knowledge?', an enumeration of cases of knowledge, and 
the commentators (adds Bambrough) support him. But * when Socrates asks 
for a definition instead of mere examples, and Theaetetus asks for an  
explanation of what a definition is, Socrates does not define definition: he gives 
examples. And now the commentators are silent.' (Italics are Bambrough's.) 
I would rather not be completely silent, for this does not seem to me quite 
fair. To begin with, and most important, Theaetetus has not asked Socrates 
to define a definition. On the contrary he says that he understands and agrees 
with Socrates's demand for universal definitions and claims to have produced 
them in his own subject, mathematics (i47C-48b). It is only in the special 
case of knowledge that he has experienced difficulty, though he has given 
much attention to it and cannot get the question out of his mind. He has no 
doubts as to the legitimacy of the question itself. Mr Bambrough, agreeing 
that it was not quite fair to cite Theaetetus in this connexion, has suggested to 
me that Wittgenstein would have done better to refer to Meno or Laches. 
(On the usefulness of definition in Laches see vol. iv, 244 f.) That he does 
seek to make his point through Theaetetus is surely quite a serious fault. 
Next, though in the particular case of the Theaetetus Socrates simply gives 
an example sufficient to remind Theaetetus, who has heard reports of the sort 
of questions he asks, one may make two further points. A) This is not true as 
a generalization about him. At Gorg. 463 c he refuses to say whether rhetoric 
is good or bad * until I have answered the question what it is', and this he 
proceeds to do. B) He has in many of Plato's dialogues, both early and late, 
answered quite explicitly the specific question what a definition is. The  
following sentence does not occur totidem verbis in any one passage, but adds 
nothing to what Plato has repeatedly affirmed: ??????? ???? ????? ??? ??? 
??????????? ????????? ?????? [or ???. . . ??????], ???????? ???? ???? ?????? 
??? ??????? ??????? ?????????? ?????? ????? ??? ???????? ???? ??? ?????? 
??? ????? ?????. (See Meno passim, Pol. 258 c and many other places. I 
am not saying here that the Platonic Socrates was right to offer universal 
definitions, only that he sometimes does so.I 
Plato himself has anticipated his modern critics by saying later in the 
Theaetetus that the idea of knowledge as true judgement plus a ????? is 
unsatisfactory if ????? expresses a mark or sign (????) by which the object 
of enquiry differs from everything else, because one cannot state such a ????? 
unless one already knows what knowledge is, and is bound to commit the 
error of including the definiendum in the definition Bo8c-ioa). I doubt in 
1 Cf. the quotation from Russell in vol. iv B44 n. 1) on the advisability of defining wisdom 
before taking practical steps in connexion with it. 
70 
Theaetetus 
any case if the Platonic Socrates ever spoke of definition as a means to 
knowledge, rather than as evidence that one already possessed it. 
Plan of the enquiry. In spite of some digressions, the Theaetetus pursues 
a more orderly and systematic course than many dialogues, and this 
may be briefly outlined without forgetting Stoelzel's warning that 'it is 
immensely difficult to force a work of art like the living Platonic 
dialogue into a rigid schematic arrangement'. 
When they have settled the difference between a string of instances 
and a universal definition, and Socrates has explained his art of  
intellectual midwifery, three suggested definitions of knowledge are tested 
and found wanting. 
i. Knowledge is sense-perception (i5id-86e). This is rejected on the 
ground that to grasp a thing's being or essence, as well as concepts such 
as similarity and dissimilarity, good and evil and the like, the mind must 
go beyond sensation and use its peculiar powers of reason and  
reflection. Without Being, no one can reach the truth, and a man who cannot 
reach the truth cannot be said to know. 
The section includes a discussion of two particular theories which 
according to Socrates are closely related and both imply that sensation 
is the same as knowledge: 
(i) Protagoras's 'man the measure' theory of knowledge. 
(ii) A remarkable theory of perception based on the view of extreme 
Heracliteans that the only reality is perpetual motion, change, process. 
Interlude A72C-77C). A reminder that, whatever the conclusions of 
this dialogue, the philosopher knows of another world than this, one 
purged of evil, to which he may aspire through 'imitation of God' in 
righteousness and wisdom. 'But', Socrates abruptly concludes, 'that is 
not our present business.' 
2. Knowledge is true belief (iSjb-201 c). Rejected very briefly on the 
grounds that one can have a true belief without knowledge, as when the 
account of a witness rightly persuades one of its truth, though only  
personal experience of the events described can properly be called  
knowledge. 
In proposing this definition, Theaetetus says that he cannot simply 
suggest 'belief, because there is false belief as well as true. This leads 
71 
Parmenidesy Theaetetus^ Sophist^ Politicus 
at once to a discussion of how false belief, or error, is possible, which 
occupies most of the section ostensibly devoted to true belief. 
3. Knowledge is true belief plus an account, with the corollary that only 
that can be known of which an account or description {logos) can be 
given Boic-iob). 
This raises the question what can be the subject of a logos, and they 
examine a theory that only compounds can be described, whereas their 
simple elements can be neither described nor known but only  
perceived and named. However, reason and experience, it appears, both 
show that a compound cannot be more knowable than its elements or 
parts. 
Three possible senses of logos are then discussed, to see if they could 
turn true belief into knowledge: 
(i) Speech in general, the expression of thought in words. This is 
dismissed as much too general. 
(ii) Enumeration of parts or elements. But addition of this to true 
belief will not give the guarantee of future correctness which is  
demanded of knowledge. 
(iii) Ability to name a mark by which the thing in question differs 
from all other things. This proves to be circular, for it amounts to  
saying that knowledge is true belief plus knowledge of what makes the 
object unique. 
So we end. The dialogue has not achieved its object, but has not (says 
Socrates) been fruitless, for if Theaetetus should have other  
brainchildren they will be the better for the present scrutiny, and if not, he 
himself will be a better and more amiable man for no longer thinking he 
knows what he does not. 
Introductory conversation. With his usual skill, Plato leads us gently into 
the discussion by letting the speakers make themselves known. 
Theodorus, 'expert in geometry, astronomy, calculation, and a man of 
general culture* A45a), introduces Theaetetus to Socrates as a youth 
of quite exceptional intellect and character, and moreover one who 
resembles Socrates in physical features. Typically, Plato adds the purely 
personal detail that his father left him a fortune but it has been 
squandered by trustees. Socrates quickly puts his question about the 
72 
Theaetetus 
nature of knowledge, and Theaetetus reveals his own bent when, to 
show that he has grasped the difference between exemplification and 
universal definition, he illustrates it by a point in mathematics that had 
occurred to himself andta fellow-student, namely that the geometrical 
equivalents of what are now called surds could be grouped in one class 
and given a single name (' powers') by virtue of their common character 
of irrationality or incommensurability.1 He cannot as yet, however, find 
a similarly universal formula to cover the different kinds of knowledge 
or skill, though the question is always in his mind. This shows that his 
mind is pregnant, indeed in labour, with some offspring, and needs the 
aid of that mental midwifery which Socrates, though barren of  
knowledge himself, knows how to practise on others.2 Let Theaetetus only 
say boldly whatever he can, and Socrates will assist him, not least in 
judging whether his idea be a proper child or a changeling. 
(i) Knowledge as perception A51 d-86e) 
Thus emboldened, Theaetetus replies that, as far as he can see at 
present, knowledge is nothing but perception {aisthesis): whoever 
knows something is perceiving it. 
1 For literature on the mathematics involved in Tht.'s example see Friedlander hi, 488 n. 16. 
Cf. also M. Brown,' Plato disapproves' etc., P.'s Meno 236ff. It is simply explained by McDowell, 
Tht. 116. The general definition arrived at is, in effect, 'a power is the square root of a non-square 
integer'. It is a good illustration of how mathematics can be propaedeutic to philosophy, as P. 
teaches in Rep. 7. 
* On S.'s midwifery see vol. hi, 444 f., 378 n. 1. P. may have connected it with anamnesis, the 
ideas brought to birth by S. being innate (Cornford, PTK 27k), but cf. Hackforth, Mnem. 1957, 
I28f. The passage contains a spirited defence of S.'s annoying habit of continually questioning 
while refusing to give his own opinion. Cf. esp. ijob-c, 151C5-8, Rep. 336c, 337a. The possible 
connexion with anamnesis is denied by McDowell (Tht. p. 117) on two grounds which make one, 
if anything, more inclined to believe in it than before. He says : A) 'The "offspring" delivered 
by Socrates are just as likely to be incorrect as correct'; B) 'Second, the Theory of Recollection 
contains nothing corresponding to the barrenness of Socrates himself.' Comment: A) In Menoy 
the locus classicus for the theory of recollection, the slave does give several incorrect answers. 
True, each wrong answer brings him nearer the truth, but so do the formulation and discarding 
of three wrong answers in Tht. (See 210c.) B) In the Meno too, S. says that Meno's question 'Is 
virtue teachable?' implies the prior question ' What is virtue?', and that to that he does not know 
the answer G1 b). Anamnesis in fact provides the solution to the problem of how one can look for 
something that one does not know. How this would work out in terms of a metaphor of pregnancy 
and giving birth, it might be hard to say, and since that metaphor is not used in Meno, it would 
be wrong to try. 
73 
Parmenides, Theaetetus, Sophist, Politicus 
What is included in aisthesis {usually in this chapter 'perception l)? It was 
a wide term, for the Greeks had no single words to distinguish sensation 
from perception, i.e. mere awareness of sense-data (colours, sounds 
etc.) from the perception of external objects which derives from it. 
It was used of a single sense, and in the plural of the five senses, but 
also much more widely, as when Thucydides B, 50) says that the dogs 
of Athens, by refusing to touch the corpses of those who had died in 
the plague, provided the best aisthesis (visible evidence) of its effect on 
animals. Plato does not confine it to a single technical use. Even when 
using it narrowly, he includes pleasures, pains, desires and fears along 
with sight, hearing and smell as aistheseis* adding that there are 
innumerable others, named and unnamed. But he can draw the  
distinction when he wishes, even inventing a word (???????? 182 a) to  
distinguish a sensible quality (or sense-datum) from the object that it 
qualifies, e.g. whiteness from a white stone (i56e5).3 In general  
however what we receive through the senses includes any direct or unbidden 
experience, as distinct from the results of rational reflection;4 and the 
latter are confined to highly abstract concepts like existence, unity, 
sameness, dissimilarity and their opposites A85 c). Things white, hot, 
sweet, or hard the mind perceives through the senses, and though the 
neuter plurals point to things rather than qualities, it is extremely  
doubtful whether at this point Plato had the distinction in mind. (See i84c-e; 
but' sound and colour' at 185 a.) Fine distinctions are well enough when 
they affect the immediate argument. Otherwise, to depart from the 
1 A word ' now normally restricted to sense-perception—to the discovery, by means of the 
senses, of the existence and properties of the external world' (Hirst in Ency. Phil, vi, 79). 
2 156b. For P.'s views on the status of pleasure and pain see Tim. 646-6^ b: pain results from 
a sudden and violent disturbance of the bodily condition, pleasure from its sudden restoration to 
normal. 
3 Nakhnikian (R. of Metaph. 155-6, 129 k) usefully draws attention to these passages, and 
mentions as a second criterion for distinguishing them that sensation is private and irrefutable, 
perceptual reports are public, objective and testable. Together, he considers, they justify him in 
treating the theories of sensation and perception separately in discussing the whole of 151 d-86e. 
Cooper's article 'P. on Sense Perception and Knowledge' (Phron. 1974) is an interesting  
discussion of this subject from a modern standpoint. 
4 At 185 a-b not only ?????? ? but also ??????? and ??????????? and a 'that' clause ('that 
they are gods') are used of perception (???????). This might have given more support to 
Gulley's claim (?????) that ??????? includes ???? (judgement or belief, clearly distinguished 
from it in other dialogues) than the passages he actually refers to. Of these, 161 d speaks of'what 
a man judges by means of sensation' (cf. the distinction between ? and ??' ?? at 184 c), and 179 c 
actually distinguishes ????????* from judgements based on them. 
74 
Theaetetus 
usages of ordinary Greek is ungentlemanly pedantry A84c). Thus at 
Rep. 608 d Socrates asks, ' Have you not perceived (???????) that the 
soul is immortal?' One must also remember Plato's dichotomy of 
everything conceivable into sensibles and intelligibles {aistheta and 
noeta). The latter are eternal, invisible and wholly real. Aistheta include 
our whole world and whatever happens in it, the whole realm of 
Becoming as opposed to Being. 
Protagoras and his 'secret doctrine' A51 e—55 d). Socrates immediately 
says that Theaetetus's suggestion is identical with the implications of 
Protagoras's famous dictum that 'man is the measure of all things',1 by 
which he meant that everything is for any individual exactly what he 
perceives it to be, so perception is always infallible—is knowledge in 
fact.2 This was doubtless based on a theory of perception held by the 
great man as a secret to be divulged only to his pupils. (The irony is 
obvious, especially when one remembers that Protagoras only took 
paying pupils.) It holds that all things are in continuous movement and 
mutual mixture, to which they owe what is wrongly called their 
existence. Nothing should be spoken of as being, either absolutely ('in 
and by itself) or in the sense of having a definite property,3 being large 
or small or white, but everything is becoming, a product of flux and 
motion. Motion is the universal creator and sustainer: life, like fire, 
comes from friction, and depends on the movement of the sun, exercise 
preserves the body, processes of learning and practice the mind. Motion, 
as preservative, is good, stagnation destructive and bad.4 What we call 
colour is not a separate thing, whether inside or outside our eyes, but 
arises from the meeting of our eyes with the appropriate motion. It is 
1 152c He 'put the very same thing in another way'. Yet it appears later that he did not 
confine knowledge to perception. See p. 86 n. 3 below. 
2 152c5-6. Cf. McDowell's trans. I see no reason to adopt White's desperate expedient of 
excising ?$ ???????? ???? (Phron. 1972). A full account of Prot.'s doctrine will be found in vol. 
in, 171-5 and 183-92. 
3 152d. Cf. Crat. 439d: on the flux-theory one cannot say of anything that it is either 'that' 
or 'suchlike' (????????). 
4 At 152? this theory is sweepingly ascribed not only to Prot. but to all previous thinkers 
except Parmenides. Though the Ionians and Empedocles taught of generation through mixture 
of opposites, P. is obviously thinking primarily of the flux-doctrine of Heraclitus, as suggested 
e.g. by the association of fire and life (i53a-b) and the idea of motion as not only universal but 
good. (Cf. vol. 1, 454, 462, and Heraclitus A 22 DK.) 
75 
Parmenides, Theaetetus, Sophist, Politicus 
neither what meets nor what is met, but an event occurring between 
them, peculiar to each percipient. We cannot affirm that a colour 
appears the same to us as to another man or animal, or even to our 
changing selves; and this could not be so if what we perceived were 
itself large or white or hot. 
There follow A54b—55d) certain paradoxes concerning relative size 
and number which have been thought both unreal and irrelevant,1 and 
Theaetetus himself cannot see their connexion with the present topic 
A55 d). Six dice are more than four, but put them beside twelve and 
they will be fewer. Socrates is taller than Theaetetus, but when  
Theaetetus grows he will be shorter. But how can fewer become more, or a 
taller man become shorter, without changing his size? This question of 
relational properties was explained in the Phaedo by reference to the 
Forms (i02b-i03a), but from the Parmenides A31 c-e, p. 41 above) it 
could appear that Plato saw difficulties in that, as well he might. Here 
Socrates himself enlarges on the difficulties: the question seems to hint 
at inconsistency2 in three statements on which they both agree, namely 
A) nothing becomes greater or less in size or number so long as it 
remains equal to itself; B) whatever has nothing added to or taken from 
it remains equal (the same in amount); C) what formerly was not  
cannot later be without a process of becoming. To find the connexion of 
all this with Protagoras, Socrates proposes to examine the mysteries of 
certain unnamed Cleverer* or Subtler* thinkers.3 
1 So Cornford (PTK 41), and Russell (quoted by Brown, see below) dismissed the trouble as 
'an infantile disease of philosophy'. Its relevance to the argument has been defended by Bluck 
(PCPS 1961, 7-9), and M. Brown explains the point as a mathematical one (JHP 1969, 373 ff.). 
S. does not simply say that 6 is greater than 4 and less than 12; it exceeds 4 by a half and falls 
short of 12 being half, i.e. the difference between 6 and 4 is the same fraction of 4 as the difference 
between 6 and 12 is of 12 (harmonic mean). This links up with Tht.'s work on irrationals, and 
Brown sees P. as influenced by these mathematical advances and problems to see difficulties in his 
own epistemology as hitherto conceived. For some earlier opinions on the significance of the 
puzzles for P.'s thought see Cornford, PTK 43-5; Ross, PTI102; Runciman, PLE 18. 
* On ??????? ???? ?????* see Hackforth in Mnem. 1957, i3of. 
3 Actually the puzzles about relative predicates are never returned to, nor is it absolutely clear 
that 'S. promises that the theory of perception he is about to expound will contain the solution' 
to them. So McDowell (Tht. 135), but I am not entirely happy about either his or Cornford's 
translation. More literally S. says (i55d5-ei): 'Do you understand why these things being as 
they are follows from the doctrine of Protagoras?' (Tht.: 'Not yet'.) 'Then you will be grateful 
to me if I examine with you the hidden truth of the thought of a famous man—or rather, famous 
men.' He does not explicitly promise a solution, and we are left to infer that in a world where all 
is change and becoming, the problem of something being now small, now large, without an 
intermediate process of becoming, loses its meaning. 
76 
Theaetetus 
The cleverer theory of sensation (i56a~57c). This is in fact a refinement 
on the Secret doctrine' of Protagpras, and together they present an 
astonishingly advanced and imaginative theory. Two accounts are 
given, because Theaetetus does not follow the first, and it is a pity 
that Socrates did not start with the second, fuller one, as they are 
not in every detail easy to reconcile. In the first, all is motion, but 
motion is of two sorts, active and passive. From intercourse and 
friction with each other these two motions (i.e. sensible object and 
sense-organ) give birth to twins, an act of perception and a percept 
(colour, sound etc.). The second account is in several stages, (i) 
Motions are now re-divided into quick and slow. Both motions in the 
previous account are slow, and move always in the same place.1 (ii) 
When one of them—e.g. an eye—and an object structurally adjusted 
to it2 come near enough, they engender the quick motions which 
traverse space, i.e. a colour and the sensation of that colour, unique 
to the particular pair that engendered them. Then, as (iii) 'vision's 
from the eyes and colour from the other parent traverse the space 
between,4 (iv) the eye becomes filled with vision and sees, becoming 
not sight but a seeing eye, and its partner is suffused with colour and 
becomes not colour but coloured, whether stick or stone or anything 
else. 
All other sensations work in the same way.5 They have no being of 
their own, but arise from intercourse and motion: nothing is an agent 
until it meets a patient, and what is agent in one encounter may be 
1 Of course an eye or ear, or skin sensitive to touch, moves around, but only as Aristotle 
would say per accidens, because moved by the person, not by any motion of its own. Its proper 
motion is alteration, included in ?????? at i8id5· 
* ?????????, lit. 'commensurate'. The terminology is from Empedocles's theory of sensation 
by pores and effluences, for which see vol. n, 231 f., 234-7 (sight). 
3 Presumably a sort of ray, or Empedoclean effluence. Cf. Tim. 67 c. 
4 This hardly seems to fit the immediately preceding description, though no commentator that 
I know seems worried about it. There the active and passive motions must be in actual contact 
and generate quasi-sexually by friction between them. Here however the slow motions (which 
one must assume to be the same things, namely percept and percipient) need only 'approach' 
(??????^???, 'come within range' Cornford, adopted by Nakhnikian; a very different metaphor 
from Plato's of copulation!), and they give birth although a space remains between them into 
which the offspring are projected. Crombie says without apparent unease ' that when contact is 
established between subject and object, a twin progeny is begotten . . . and that these travel 
between the two parties' (EPD 11, 10 my italics; cf. p. 7). 
5 15667. P. does not explain how this mechanism can account for desires and fears. For 
guesses see McDowell i37f. 
77 
ParmenideS) Theaetetus, Sophist, Politicus 
patient in another.1 Nothing ?, everything is in process of change, and 
ideally the verb 'to be' should be excised from our vocabulary, along 
with other Static' words like 'this' or Something'.2 
The theory seems to borrow features from both Heraclitean flux and 
the atomism of Democritus. He too taught that sensations are a 
momentary product of physical contact, that sensible qualities have no 
independent character {physis), sensation being a result of alteration in 
our bodies caused by the impact of a stream of atoms thrown off by the 
object perceived, and moreover that our bodily conditions are in any 
case changing through age or other causes, so that the effect will be 
different not only on different people but on the same person at different 
times. (Cf. Thu 159 b—d.) He too quoted sweet and bitter as examples 
of the relativity of sensation. He even added the refinement that in the 
case of sight the atoms from the object do not enter the eye directly, but 
meet effluences from the eye itself and form jointly with them an image 
which does enter it. We need not deny a debt here because Socrates 
distinguishes the theory's authors, as 'much subtler', from those who 
' only admit the reality of what they can get a grip of with their hands, 
not of actions or comings-to-be or anything invisible' A55 c). These, 
commentators argue, are the materialists, and Democritus was a 
materialist. Nevertheless he did posit continuous motion and did deny 
reality to all sensible qualities, the only realities being atoms and void, 
neither of which could be seen or grasped with the hand. Even if he 
'made all sense-objects tangible', he did so only in the sense that 'most 
of the natural philosophers' did (Arist. De sensu 442a2C)-bi), and this 
did not give the sense of touch any advantage over the others. It 
appears with them in fr. 11 as one of the untrustworthy, 'bastard' 
faculties.3 
1 'The eyeball can be seen by another eye, the flesh touched, etc.' (Cornford). S. gives no 
examples. 
2 Is 157b (habit makes us use these words, though wrongly) another reminder of Empedocles? 
Cf. fr. 9.5 (on ????????) ? ????? ?? ???????? ???? ?' ???????? ??? ?????. 
3 For another view see Campbell, Thu xli-lv, and for a detailed account of D.'s theories of 
perception vol. n, 438-49. The theory of the ?????????? has been attributed in modern times, not 
very compellingly, to Antisthenes and Aristippus. (For some refT. see Friedlander, PL in, 488 
n. 20.) It is also held that P. himself either constructed the theory or at least believed it. So 
McDowell, Thu 130, preceded by Friedlander, Cornford, Jackson, Burnet, Stenzel, Ritter, 
Nakhnikian and others: contra, Taylor, PMW 329f. Runciman (PLE 19) argued that it could 
not be P.'s because he never held a Berkeleian theory of sensation, which would have conflicted 
78 
Theaetetus 
In summing up the theory to test Theaetetus's assent, Socrates says 
at 157c!: 'Tell me whether you like the idea that nothing is good or 
beautiful or all the things we have just spoken of, but all are becoming.' 
The sudden introduction of'good' and 'beautiful' into what had been 
a list of sense-perceived properties like white and hot may sound odd, 
but for Plato all alike belong to the sensible world. What he has in mind 
is the 'many beautiful things' of Rep. 5, which are recognized by the 
' lover of sights and sounds', and in fact are no more beautiful than ugly, 
in just the same way as large and small, heavy and light things (both in 
the Republic and here) can appear as their opposites. There too, exactly 
as here, he says that none of the many phenomena are, rather than are 
not, what they are said to be. (See Rep. 479 a-b.) 
Status of the sensible world. A point vital to Plato's philosophy arises 
here. It is pressed home at i82cff.: on the theory that everything is in 
unceasing change (flux) we cannot even say that a thing 'flows white', 
for the whiteness itself is flowing and shifting into another colour. We 
cannot name anything with any assurance that we are naming it rightly, 
or even say that perception is knowledge any more than non-knowledge. 
In short, the theory makes all discourse impossible. 
Now in the Phaedo and Republic Plato teaches that sensibles are 
always changing but at the same time can 'remind' us of the changeless 
Forms—the only realities—because they resemble them or in an  
imperfect and timebound manner 'share' their natures. But if we assume (as 
many do) that Plato accepts for the sensible world the extreme form of 
the flux-doctrine which we have here, then, as Gulley writes (PTKJ4), 
What becomes of the doctrine that sensible characteristics are * copies' or 
* images' of Forms, that they are recognisable and hence are able to prompt 
the recollection of Forms? This doctrine clearly assumes that there are 
determinate and recognisable sensible characteristics; indeed it is a doctrine 
that sensibles are determinate and recognisable in so far as they * participate 
in* and hence * resemble' Forms. There is a serious inconsistency, then, 
between this doctrine and the consequences drawn by Plato from the fact 
that sensibles are in flux. 
with the theory of Forms, but Cornford (PTK 50f.) seems to have thought it not exactly 
Berkeleian. (For comparison and contrast with the Berkeleian phenomenalist tradition see 
McDowell 143 f.) In fact it could not be P.'s for the reasons given on the next few pages, and is 
plainly the neo-Heraclitean doctrine referred to as such at I79dff. 
79 
Parmenides, Theaetetus, Sophist, Politicus 
Others have expressed similar views,1 but the point is, I suggest, that 
Plato here describes the sensible world as it would be if there were no 
Forms. Neither supporters nor opponents of this explanation have 
appreciated that their existence changes the nature of the sensible world. 
This indeed was a m^in reason for their introduction. Parmenides had 
denied all reality to the sensible world on the ground of his exclusive 
dichotomy 'is or is not'. The Forms, and the admission of becoming' 
as an intermediate stage, were designed, not to depress the sensible 
world but to save it from annihilation. Somehow the Heraclitean and 
Parmenidean views of reality must be reconciled. The Platonic universe 
is an integrated whole consisting of intelligible and sensible spheres. As 
the Timaeus teaches—that triumphant vindication of order, regularity 
and value in the movement and change of the sensible world—what gives 
it such order and stability as it possesses is the fact that it is modelled on 
the Forms.2 It is true that for Plato Sensible things are forever flowing, 
and there can be no knowledge of them'; but there can be true opinion 
because, Aristotle continues, there are also what he called Forms, with 
reference to which the sensibles can be spoken of because Forms are 
their causes, that is, they impart definite characters to the sensibles.3 
Contemporary Heracliteans were like their master without his Logos, 
the universal law governing the continual flux of change,4 and Plato 
had no thought of following them in their fantasy of a world adrift on a 
sea of indescribability. We may recall the outburst against them of 
Theodorus, an authoritative and sympathetic voice in the dialogue: you 
might as well, he says, talk to maniacs; they are living examples of their 
theories, always in motion, incapable of staying still a moment to listen 
to a question or an argument. They own no masters or pupils, it's a case 
1 Cornford held that the extreme flux-doctrine was P.'s own theory of the sensible world: 
having proved by its means that knowledge cannot be perception, he leaves us to infer that it 
depends on the Forms. This would certainly make him guilty of the inconsistency which Gulley 
finds. If Forms existed, yet in no way moderated the utter instability and disorder of our world 
(an inconceivable situation), knowledge would be as impossible as if they did not. 
* Cf. esp. 52a. On the 'cleverer' theory, of course, one could not speak of a sensible as 
coming to be 'in a certain place' Bv ???? ????). Cf. also Gorg. 507e~5o8a. 
3 Arist. Metaph. 987332^9. The causal aspect of the Forms has been prominent in many 
places in dialogues already discussed. See on it vol. iv, 350-2. 
4 For a summary of H.'s conception of the Logos see vol. 1, 434. His distance from his rabid 
followers is also indicated by fr. 55, where he makes a point similar to that made later (i86d) 
by P. himself, that the senses are 'bad witnesses' which cannot yield knowledge without a mind 
to interpret them. 
80 
Theaetetus 
of spontaneous generation, and each thinks the other an ignoramus 
(i79e-8oc). Theirs was the position of Cratylus, who outdid Heraclitus 
by saying that one could not step into the same river once, and ended by 
taking Plato's hint and abandoning speech altogether (Arist. Metaph. 
ioioai2-i5). 
At the end of the Cratylus D39 b ff.) Socrates demonstrates to Cratylus 
that on his extreme flux-theory, allowing no permanent entities at all, 
verbal communication and knowledge would be impossible. If not only 
beautiful things but the very property of beauty were under constant 
change, there would be nothing to which one could apply noun or 
adjective, as having either identity or qualities. This does not show that 
Plato 'does accept that the sensible world is in flux* (in the extreme 
sense) and so * at the same time he asserts that Forms exist and denies 
that the sensible world has any determinate characteristics. This* 
(Gulley goes on to claim) * is in itself implicitly to acknowledge that the 
argument that "being in flux" is incompatible with "being  
determinate " is equally valid whether or not it is assumed that Forms exist. '* 
All it shows is that the neo-Heraclitean theory that everything is in 
incessant flux and change is inconsistent with the existence and effect of 
the Forms. Finally, in the Timaeus, in which the Forms are assumed 
from first to last, it is said, in contrast to Crat. 439 d 9, that whereas 
physical bodies such as fire and water, being mutable and unstable, 
cannot be called * this' or 'that' (?????), they can be said to possess 
certain qualities (????????) by virtue of the penetration into their 
habitat of copies of the Forms.2 In their causal capacity the Forms 
1 Gulley, PTK 83. On p. 72 he admits that this thesis involves rejecting the 'grammatically 
more obvious interpretation' of 439c!4. Even if (which I do not believe) P., never a precise 
writer, had been betrayed into giving the impression that the existence of the Forms made no 
difference to the nature of the sensible world itself, the weight of evidence on the other side 
would far outweigh it. Runciman saw the point. See his PLE 21 on the argument of the Cratylus. 
On G.'s views see also vol. iv, 493 n. 1. I believe that what I say here is also relevant to 
the remarks of Robinson in Essays 48, and Owen in SPM 323. 
2 Tim. 49 d, 50 c. This should be read in conjunction with Cherniss's acute arguments in 
SPM 355-60, though I do not necessarily follow him in all their subtleties. The late dialogue 
Phil. E9a-b), though it puts the contrast between being and becoming in strong terms, only 
repeats the point made in Rep. 5 that precise truth cannot be found within the changing sensible 
world, and therefore ???? is different from ????????. I hope the last few pages answer the point 
raised by McDowell, pp. 180-1, para, (ii), and I believe I am in substantial agreement with the 
extremely close-knit argument of Cherry in Apeiron 1967. 
7 
8l 
GHG 
Parmenides, Theaetetus, Sophist, Politicus 
rescue the sensible world from the meaningless chaos to which the neo- 
Heraclitean maniacs would consign it. Truly, as the Parmenides repeats 
A35 b-c), if you deny the existence of Forms you will have nothing on 
which to fix your mind and will destroy the possibility of rational 
discourse. To suggest that their existence and presence could leave the 
flux of becoming unaffected shows a fundamental misunderstanding of 
Plato's position. 
Dreams and hallucinations A57e-6od). This theory (continues Socrates) 
can withstand the objection commonly made, that in dreams, madness 
or illness we have false perceptions, so that perception is not infallible 
after all. We cannot even give certain proof that we are not at this 
moment asleep and dreaming our conversation,1 and the answer to the 
objection lies in the theory's assertion that sensation is nothing more 
than an interaction between two constantly changing things, and exists 
(or rather * becomes') only in relation to both. One must always add 
the Protagorean * for him', * for me'. There is then no such thing as an 
illusory sensation. If wine sweet and pleasant to a healthy man tastes 
sour to the same man in sickness, the explanation in terms of the theory 
is that he has become a different subject, whjjh together with the 
drinking of the wine produces different offspring, namely the sensation 
of sourness on his tongue and a * moving and changing sourness' in the 
wine A59 c), which has no qualities 'in itself but only * for somebody'. 
Thus Protagoras is vindicated, and each man is the sole judge of what is 
for him; and the name of knowledge cannot be denied to a state of mind 
impervious to falsehood or error about what is or becomes. 
1 This was asserted independently by Descartes {Meditation 1, trans. Haldane and Ross 
pp. 75 f.). For Moore's and Russell's positions see Newell, Concept of Phil. 56-8. J. L. Austin 
(S. and S. 49 n. 1) says it is absurd because (for one thing) we describe some waking experiences 
as 'dream-like', and if Descartes (and P. whom neither mentions) were right, 'if dreams were not 
"qualitatively" different from waking experience, then every waking experience would be like a 
dream'. I do not believe P. was right, but I doubt if it is possible to refute him so easily. An 
experience which we call dreamlike is one which we believe to be real (not imaginary like a 
dream), but which gives an impression of the unreality which, in our waking hours, we ascribe to 
our dreams. While we are dreaming, our dream-experiences seem real (witness the way we may 
wake up laughing, crying or in a state of fear), and it is by no means inconceivable that in a 
dream we might speak of our experiences as dreamlike though (like the man awake) we believed 
them to be real. Somewhat similar is Tht.'s point that we can dream we are narrating a dream, a 
thing which I myself have often done. P. too could speak of a waking experience as dreamlike 
{Meno 85 c). 
82 
Theaetetus 
Examination of the theory that knowledge is perception. Thus Theaetetus's 
firstborn has been delivered after a difficult labour. The next job is to 
examine the baby and see if it is worth rearing. I have said that the 
Theaetetus pursues on the whole a systematic course, but it preserves 
the natural turns of a genuine conversation, with short interludes, a 
longer digression, and shifts from one aspect of the subject to another 
and back again. This realistic style is particularly marked in the next 
few sections. 
(i) Return to Protagoras (i6ib-i62a). If knowledge is perception 
and every man has his private and unassailable truth (and on this 
supposition why confine it to man among sentient creatures?), what 
right had Protagoras to set himself up as a teacher? Can he have 
seriously meant that no man is wiser than another, or even than a pig or 
tadpole? Having said this, Socrates immediately turns round and 
denounces it in Protagoras's name as cheap rhetoric. Without refuting 
it, he insists they must attack the question in a different way, and passes 
to a new point. 
(ii) Foreign languages and unlearned letters A63 b-d). Assuming that 
knowledge is sense-perception, what happens in the case of an unknown 
language? Do we not hear what is said, or do we both hear and know 
it? And again, before learning to read do we not see letters, or do we see 
and therefore know them? Theaetetus replies judiciously that we know 
just as much as we see or hear, the sound of the voices and the colour 
and shape of the letters; but we neither perceive nor hear what an 
interpreter or schoolmaster could tell us. Socrates congratulates him on 
this piece of clear thinking, which he will not dispute for fear of  
stunting his growth. He could of course have replied that to admit that 
language, besides its audible or visible symbols, has a meaning which an 
interpreter or teacher could convey, is to admit that perception is not 
the whole of knowledge. But this coup de grace—the indispensability of 
mind and ratiocination in the acquisition of knowledge—is not to be 
administered until much later (i84b-86d), to allow for further criticism 
of both Protagoras and the flux-theory. 
(iii) Memory A63 d-o^).1 Knowledge, we say, is perception. Then 
1 Gulley (PTK 77) refers to this passage as evidence that in the claim of perception to be 
knowledge, perception is meant to include memory-images. This, surely, would reduce the 
83 
7-2 
Parmenides, Theaetetus, Sophist, Politicus 
he who, for example,1 sees something knows that thing as long as he 
sees it. Are we then to say that when he goes away or shuts his eyes, he 
necessarily forgets it, or alternatively that though he still remembers 
it clearly, he no longer knows it because he does not see it? Either 
alternative strikes Theaetetus as Monstrous', and he admits that the 
limitation of knowledge to sensation apparently leads to impossible 
consequences. 
(iv) The ?knowing and not-knowing' dilemma A65 b-d). This sounds 
final, and Socrates's next move is surprising. Without refuting the last 
argument, he declares that Protagoras would have put up a better fight 
for his theory. They have been quibbling like contentious Sophists, not 
true philosophers. He must try to come to the aid of the dead Protagoras. 
But far from defending the thesis that knowledge is perception (which 
he has all along said is included in Protagoras's), what he does first is to 
attack it once more by an ultra-sophistical argument. If you look at 
something with one eye closed, do you see (and therefore know) it or 
not? Answer yes or no—no nonsense about seeing with one eye but 
not with the other. Under such pressure Theaetetus agrees that the 
only possible answer makes his thesis self-contradictory. Many other 
questions might be asked, continues Socrates: for instance, can  
knowledge be keen or dim (like perception)? Then, throwing this argument 
aside, he goes straight on to ask how Protagoras would defend his 
position. 
The argument is identical with some of those used by the fighting 
brothers in Plato's farcical exposure of eristic in the Euthydemus. It 
depends on demanding a simple answer in the terms of a question using 
an incomplete predicate or in some other way unanswerable without 
argument to nonsense. Cornford {PTK 65) says S. breaks it off because to save the definition of 
knowledge as perception that term must be stretched to include memory (true enough), and 
* there would be no objection to that*. For all P.'s variations on the scope of ???????, I think he 
would strongly object to calling memory a sensation or emotion A56b), or anything else but an 
act of the mind. In terms of the modern distinction between potential and actual remembering 
(for which see Broad, Mind and its Place in Nature 222, Shoemaker in Ency. Phil, v, 271), 
according to which one can say that a man remembers an event in his childhood even when he 
happens to be asleep or thinking of something else, P. seems to be considering memory-acts only, 
not memory-powers. 
1 Seeing is of course only one example of perception: P. could equally well have spoken of 
remembering a tune one has heard. (Cf. vol. iv, 508, n. 5). But in speaking of memory, as most 
often of knowledge in general, he has in mind acquaintance with an object or person rather than 
knowledge of a fact. 
84 
Theaetetus 
qualification.1 It cannot have been meant seriously by Plato, who 
indeed emphasizes his irony by describing its proponent as 'an  
imperturbable gentleman', 'a targeteer serving for pay in the army of 
words \2 Why should he produce this succession of arguments which 
Socrates either drops abruptly or himself dismisses as contentious? 
Because, I suggest, though he enjoys playing with the indefensible 
thesis that all knowledge is provided directly by the senses, and takes 
Protagoras seriously enough to want to examine him from every point 
of view, there is for him only one unassailable refutation of these 
theories, which he is saving for the end: the need for mind, which can 
go beyond the senses to use its peculiar power of reason, drawing its 
own conclusions from the data which the senses present but cannot 
interpret. Only mind can fulfil the essential condition of knowledge by 
reaching the essence (ousia) and the truth of things (i86c-d); and when 
ousia and truth are contrasted by Plato with sense-perceptions, this can 
only mean that the sensible world is to be interpreted and understood 
in the light of the Forms. 
(v) Back again to Protagoras: the defence (i65e-68c). Most of the 
defence is delivered by Socrates in direct speech, as from the mouth of 
Protagoras, with plenty of scolding of himself for unfair tactics.3 
' Protagoras' deals first with the last two arguments against identifying 
knowledge with sensation, and then at greater length upholds his own 
(historically genuine) views. 
(a) To the argument from memory he replies that the memory of a 
past experience is something different in kind from the original  
experience. This would meet the question whether a man remembering 
something that he has seen nevertheless does not know it: he knows the 
memory-impression but not the object of his sensation.4 (It could also 
raise the unmentioned question, what is a memory-impression if it is 
distinct from a perception and all knowledge is perception?) 
1 Examples are the familiar Parmenidean * Can a thing both be and not be?' at 293 c and * Who 
learn, the wise or the ignorant?' B75 d). See also the summary of the Euthyd., vol. iv, 268 ff., and 
276 f. 
2 This character will ask an ??????? ???????, reminding us of the boast of the brothers, 
?????.. .???????? ?????? (Euthyd. 276?). 
3 The dramatic and other significance of S.'s elaborate and entertaining impersonation of 
Protagoras has been vividly brought out in ?. ?. Lee's article, Exegesis, 225-61. 
4 For a recent discussion of this argument see E. N. Lee, I.e. 235. 
85 
Parmenides, Theaetetus, Sophist, Politicus 
(b) The question whether someone can know and not know the same 
thing is meaningless, for in a world where both subject and object of 
perception are continually changing, one cannot speak of the same 
person or the same thing at all. Socrates might have added here what he 
says later (i84d), that this theory treats a man as a collection of separate 
sense-organs, with no psyche (mind or personality) to unite them. On 
that hypothesis it could be legitimate to say that one eye sees and knows, 
the other not, but psyche is not to be mentioned yet. 
(c) ' Protagoras' now returns to his own ' man-the-measure'  
statement (which clearly does not involve confining knowledge to  
perception),1 and Socrates's ' vulgar and unscientific' objection that it would 
preclude any man from being wiser than another, or even than a beast. 
None can dispute that my beliefs are true for me, but it may be better for 
me that other things should both appear and be to me true. Just as the 
doctor with medicines alters a patient physically to give him pleasant 
sensations instead of painful (his indubitably sour wine appears and so 
becomes for him sweet again), so a Sophist can with persuasive words 
change a man mentally so that he has thoughts which, though not truer 
than formerly, are more profitable. Even the customs and laws of a 
state are always right and proper for it so long as it thinks them so, yet 
may in practical terms be harmful, and a statesman (or Sophist in his 
political capacity) can work on it by his oratory until useful and 
valuable practices both appear and are so for it. The test of truth or 
falsehood is replaced by the pragmatic one of future benefit or harm.2 
(vi) Protagoras continued: criticism of the defence (i7oa~72b). (a) 
Everyone except Protagoras thinks false beliefs possible.3 By his own 
doctrine he must concede that their belief is true for them, and it is 
therefore more false than true by as much as the rest of mankind  
outnumber his single self. 
This argument is perhaps not very serious, but at least it is not, as one 
might suspect at first sight, contradicted by Socrates's insistence  
elsewhere that truth is not to be decided by a counting of heads, and that 
for his own part if he were convinced that something was true the fact 
1 Cf. vol. in, 186 n. 2. 
2 For a full account and discussion of this curious doctrine see vol. in, 171-5, 267^ 
3 Here S. frankly carries Prot.'s doctrine beyond the field of sensation, using the words 
??????? ?, ???????, ??????? (i7od). 
86 
Theaetetus 
that no one else believed it would leave him unmoved.1 He would 
simply say that the others were wrong, but Protagoras cannot do so, 
and as a preliminary dig it is well enough to say that according to him 
there must be ??-thousandfold more truth in the denial of his doctrine 
than in its assertion. 
(b) Protagoras refutes himself. When he admits the truth of his 
opponents' contrary belief he is himself agreeing that his own is false, 
i.e. it is untrue that any man, however ignorant, is the measure of 
truth A71 a-c). This refutation was neatly summed up by Sextus, who 
attributes it also to Democritus {Math. 7.389): 'If everything that 
appears is true, the belief that not everything that appears is true, being 
based on what appears, will itself be true, and so the belief that  
everything that appears is true will become false.' 
The simple syllogism: 'Every belief is true; some men believe that 
not all beliefs are true; therefore some beliefs are false' sounds cogent, 
and strongly suggests that the dictum of Protagoras, like the paradox 
of the Liar (vol. in, 499), involves a vicious circle. In the past,  
commentators have either passed over Socrates's argument in silence (e.g. 
Campbell) or called it fair (Cornford). Lately, however,2 attention has 
been drawn to the fact that Socrates has omitted the qualification 
hitherto scrupulously inserted, that the contrary belief of others is only 
true for them. Could not Protagoras reply that his doctrine remains true 
for him, though false for others? Against this it is said that the belief of 
others is not, like his own, that the doctrine is false^r them, but that it 
is absolutely, or objectively, false, and this therefore is what he is 
acknowledging to be true.3 Most recently ?. ?. Lee has maintained that 
it is still open to him to say:' Certainly it is true for me that it is true for 
them that my view is simply false; but that is because they cling to the 
old vocabulary of objective falsehood which I have shown to be 
inadmissible. If I say their view is true for them, I am not committed to 
saying that it is true for me.>4 On the other hand, while rescuing him 
1 Gorg. 471 eff., H. Ma). 298 b. 
2 See Runciman, PLE 16 and Vlastos, introd. to Prot., xiv n. 27. 
3 The question whether Prot.'s dictum can account for second-order judgements (judgements 
of the truth or falsehood of other judgements) has been discussed by Tigner in Mnem. 1971, 
whose view is disputed by ?. ?. Lee, Exegesis 242-8. 
4 I am not quoting Lee verbatim. 
87 
Parmenides, Theaetetus, Sophist, Politicus 
from the toils of Socrates, Lee concludes that his immunity from  
refutation is bought at the price of showing that he is not really saying  
anything serious that can be significantly discussed or denied.1 
(c) Grant that the doctrine is true in the field of sensation: each man 
is sole judge of what appears to him sweet, cold, hot etc. But even our 
defence of it showed that, in the matter of what is expedient because it 
will bring future benefits, one adviser is better than another in respect of 
truth A7238). As a judge of what will be healthy for him, 'knowing 
within himself the healthy',2 one man will not be as good as another, 
nor one state in judging what actions will benefit it, even admitting that 
what it lays down as just or unjust, holy or unholy, is so for it. As for 
such moral and religious concepts, 'men do assert' that these have no 
real, fixed nature, but are only a matter of agreement. This line is taken 
by those who do not accept Protagoras's theory completely. 
The above is a paraphrase of ijidS-jibj. Unfortunately there is 
disagreement about the last sentence. First, the subject is changed from 
'the argument' (logos, sc. of the defence) to an unnamed 'they', which 
some think only a stylistic variation, others a different reference^ and 
secondly there is doubt about those who 'do not accept Protagoras's 
theory completely'. On the evidence about the Sophists presented in 
vol. in, I believe that what Plato has in mind is this. The view that 
'things just and unjust, holy and unholy' have no nature or essence 
(physis or ousia 174 b 4) of their own but are only matters of convention 
(nomos) or agreement was shared by Protagoras with the other Sophists; 
but whereas he argued that, on grounds of simple expediency,  
established laws and customs ought to be upheld, many of them saw in the 
merely conventional basis of law and current morality a reason for a 
man to flout them whenever it suited him.4 
1 L.c. 248. Cf. earlier Runciman, PLE 16: 'He can, in fact, only advance [his belief] at the 
cost of any standard by reference to which it could be demonstrated.' 
a Another instance of a current expression which may be taken as implying the full theory of 
Forms or not, according to choice. Cf. vol. iv, 222 f. 
3 Contrast Cornford, PTK 81 n. 1, with Hackforth, Mnem. 1957, 132^ 
4 See vol. in, esp. pp. 146, 268. It will be seen that I do not agree with Cornford {PTK 82) 
that in Prot.'s belief sensations, still less moral concepts, existed 'by nature*. 
88 
Theaetetus 
At this point the argument is interrupted by the famous Digression con- 
trasting the life of the philosopher with that of the lawyer and man of 
affairs. Though Plato is unlikely to have placed it there without good 
reason* it may be simpler to finish the discussion of Protagoras first. 
(vii) Final refutation of Protagoras (ijjc-j<)b). For this Socrates has 
only to elaborate a point already made. The theory that perceptions and 
experiences are unchallengeably real and true for the experiencing  
subject may well be valid for the present and past, but it fails the test of 
prediction. Judgements of expediency concern the future effect of  
present behaviour, as to which there is no disputing that one man knows 
better than others what will appear and be to them. This applies to  
experts in many arts—doctors, vintners, musicians, cooks, legislators—and 
Protagoras himself earned large fees in the sincere belief that he knew 
better than others what would appear and be to them in the future. 
This leaves the flux-philosophers, who since they confine their belief 
in the infallibility of sensations to the present, are not touched by this 
argument. Before finishing with them, we may turn to the Digression. 
Digression: the philosopher and the practical man A72C-77C) 
Summary. The pretext for this is slight, simply a remark by Theodorus 
that if the arguments look like multiplying and getting more formidable, 
after all they have plenty of time. This prompts Socrates to reflect how 
natural it is that those who spend much time in philosophy should cut 
ridiculous figures when they appear and speak in a court, and we are 
back at the Gorgias and the reproach levelled at Socrates by Callicles. 
Plato never tires of insisting that it is in fact a mark of Socrates's 
superiority. The truth is, he continues, that compared with those bred 
up in the law-courts philosophers are as free men to slaves, having 
leisure2 to pursue any subject they like for as long as they like, with the 
sole aim of reaching the truth. The lawyer by contrast is tied to a topic 
1 On the Digression as taking up on a higher, more universal level the theme of the criticisms 
of Prot., see Lee, l.c. 238-41, 354f. 
2 Schole, 'leisure*. But as used by P. and Aristotle the Greek word acquires much richer 
associations and names a typically Greek ideal, more like our 'culture*. It is not accidental that it 
has given birth to 'school* and 'scholar*. Its value and its association with philosophy and 
learning are especially emphasized by Aristotle. When he says that happiness lies in schole (EN 
H77b4, Pol. 1338a 1) he does not mean idleness. Nature herself prompts us to use leisure rightly, 
89 
Parmenides, Theaetetus, Sophist, Politicus 
imposed by a watchful opponent and to a strict time-limit. He must 
learn the arts of flattery and deceit, clever in his own estimate but in 
reality twisted and stunted in mind. 
The complementary description of the philosopher suggests a 
Platonic rather than a Socratic ideal. He is a stranger not only to the 
law-courts and Council but to the market-place and dinner-parties. 
Only his body is in the city, while his mind is abroad seeking the true 
nature of all that exists. His interest is not in the doings of men around 
him, but in what man is, and how distinguished from other beings. 
Since birth, wealth, rank and power mean nothing to him, he appears 
both arrogant and in practical affairs helpless and ignorant. The position 
is reversed if the man of affairs can be persuaded to abandon personal 
questions of who has injured whom for the question of 'justice and 
injustice in themselves, what they are', and instead of calling rich men 
and kings happy, consider the whole nature of kingship and human 
happiness. Then it is he who will make a fool of himself. 
Theodorus comments that if everyone believed this there would be 
fewer evils in the world, but Socrates replies that evils can neither vanish 
('there must always be something contrary to good'I nor have any 
place in the divine sphere, so they haunt this world 'of necessity'. 
Hence one should make all speed to fly from here to there. This is done 
by becoming as like as possible to God, the perfection of righteousness, 
making oneself'just and holy with wisdom' (or knowledge, ????????). 
To understand this is true wisdom and excellence {arete), as opposed to 
the world's conception of them. Those who aim not at being but at 
seeming wise in the eyes of the world, whether in a profession or craft or 
in politics, are low and vulgar.2 Their penalty is inescapable. Of'two 
not squandering it in play. Schole is the whole basis of life, the goal of all business, and carries 
its own happiness and pleasure within it. In eulogizing it in the Ethics as an end in itself, he like 
P. calls the life of the politician 'leisureless*. If states do not know how to live at peace, their 
lawgivers are to blame for not having educated them in the life of schole {Pol. 133439). At Pol. 
1323 b39 'a different schole* means a different branch of learning, and at I3i3b3 scholai in pi. are 
the bane of a tyrant (*societies for cultural purposes' Barker). 
1 This is not explained. At Lys. 221 b-c S. says (though how seriously seems doubtful) that 
if evil disappeared, good would lose its value. Or P. may have in mind the thought at Pho. 97 d 
that knowledge of the best involves knowledge of the worse. Similarly at Ep. 7, 344 a-b, virtue 
and vice must be learned of together. (Cf. Arist.'s oft repeated principle ??? ???????? ? ???? 
????????.) 
% The politician whose wisdom is counterfeit is ???????? and the technician ????????. The 
latter word commonly expressed an aristocratic contempt of handicrafts. Here P. seems to be 
90 
Theaetetus 
patterns established in reality itself, one of divine beatitude and one of 
godless wretchedness, their unrighteous lives assimilate them to the 
latter, and shut them out for ever from the place where no evil can come. 
Their penalty is to live for ever on earth lives such as they live now, in 
the company of others as bad as they. All this they will dismiss as 
foolishness, though when any of them have the courage to stand up to 
questions and examination, in the end they feel dissatisfied with their 
own arguments and are silenced. 
The lesson of the Digression. This is plain. The attempts to define  
knowledge in the main part of the dialogue are carried out by every means 
short of the doctrine of Forms, and end in failure. The digression assures 
us that the teaching of Phaedo and Republic, Symposium and Phaedrus 
has not been abandoned, and that a successful search for the nature of 
knowledge lies beyond Plato's self-imposed limitations here. The whole 
spirit of the Digression sets it apart from the rest, as do many details 
within it. Like the Gorgias, it not only contrasts the characters of the 
philosopher and the man of affairs but speaks of another world in which 
both get their deserts, as in Phaedo, Republic and Phaedrus. The  
practical helplessness of the philosopher and the reasons for it were enlarged 
on at Rep. 487bff., and his neglect of the trivial and passing affairs of 
men to concentrate on universal questions of the being and nature of 
things recalls 5oob-c, where the objects of his contemplation are 
described as 'fixed and immutable realities, existing according to order 
and reason', by familiarity with which he becomes, as here at 176b (and 
at Phdr. 253a, Tim. 90c; cf. Pho. 82b-c), 'like the divine so far as a 
man can'. That there is no unrighteousness in a god was stated in the 
condemnation of traditional mythology in Rep. 2 C79 b-c). The  
contrast between popular arete and arete 'with wisdom (pkronesis)' was 
drawn at Phaedo 69a-c (where the popular kind is also called 'slavish'). 
That the righteous man is happy, and the unrighteous wretched, in this 
life is asserted at Rep. 354a. The most striking use of the language of 
the middle dialogues is the mention of 'patterns fixed in reality' at 
saying that whether an art is banausic depends on whether it is pursued in the awareness of a 
higher good, a possibility which he does not rule out. (Cornford's loose translation here gives a 
false impression.) 
91 
Parmenides, Theaetetus, Sophist, Politicus 
17663, and reincarnation is clearly implied at 177 a. As a final small 
point, the politician who submits to the Socratic elenchus, and ends by 
being dissatisfied with his own statements, so that his rhetoric fades 
away and he seems a mere child A77 b), though veiled in a discreet 
plural, is surely par excellence the Alcibiades of Symp. 2i5e-i6ai. 
A critic might claim that for all their echoes of the middle dialogues, 
not all these passages, taken singly, necessarily imply the full Platonic 
theory of Forms. But some do, and I have quoted them all to emphasize 
how the other-wordly, religious spirit of the Digression transports us 
momentarily away from the prevailing analytical tenor of the Theaetetus 
to regions from which it is unthinkable that the Forms should be absent. 
The Forms, we may conclude, remain for Plato a datum, almost an 
article of faith, but he is now much concerned with problems of their 
mode of existence and their relations both with each other and with the 
sensible world, which had not occurred to him in the magnificently 
confident mood of his middle period. These new problems necessitate 
leaving them aside at times to examine other possibilities. 
Excursus: Evil and its sources1 
Evils, says Plato here A76 a), can never be abolished, nor have they any 
place in the divine world. It may be a good moment to consider what 
we have so far learned of his ideas about the nature and sources of evil, 
with perhaps a forward glance at what is still to come, bearing in mind 
that we can only pick up references to it scattered here and there 
through the dialogues. He nowhere draws the threads together in a 
systematic account, and it cannot be assumed from the start that he had 
a final solution of this intractable problem or that his views on it  
remained consistent. Some see him as gradually shifting the responsibility 
from body or matter to soul, others have detected two irreconcilable 
concepts of evil existing side by side in his system, or declared that the 
problem never seriously concerned him at all.2 
1 Only the dialogues are considered here. On Ep. 7, 344a-b, and its possible connexion with 
the * unwritten doctrine', see Kramer, Idee u. Zahl 119. For the dialogues see also the full treatment 
by Hager, Die Vernunft und das Problem des Bosen. 
2 For a summary of views and references for the debate see Cherniss in Plato 11, ed. Vlastos, 
244 n. 1, and bibliography on p. 258. His article is largely concerned with Tim. and Laws, and 
argues strongly in favour of a consistent account based on P.'s analysis of the phenomenal world 
as a moving reflection in space of immutable, non-spatial reality. 
92 
Theaetetus 
(i) Evil as a negative conception. The Republic C79 b-c) confirms that 
evil is prevalent in human affairs and cannot be attributed to God. 
' Some other explanation must be sought.' The first thought that occurs 
is that nothing in the phenomenal world can be perfect because it  
contains only imitations of the Forms, and the mutable cannot attain the 
perfection of its eternal models. Evil appears thus as something  
negative, a shortfall from perfection and, since only the Forms are fully 
existent, from complete reality. The description of the Form of Good 
in Rep. 6 explains the inseparability of being and goodness, of both of 
which it is the cause.1 This however does not give evil a wholly negative 
character, as I pointed out earlier (vol. iv, 508 n. 1). Things in the 
physical world, though short of perfection, have their functions and 
therefore their own excellence (arete), the absence of which can be an 
active power for harm. In Plato's example of a pruning-knife {Rep. 
353 e), lack of its proper excellence—sharpness—is a positive evil which 
can harm the vine.2 Similarly man's moral evil, though a powerful force 
for wrong, results simply from a lack, for Plato never gave up the 
Socratic tenet that its source is ignorance and it is therefore involuntary. 
(It is repeated in the Laws, 731 c, 86od-6i d.) 
(ii) Evil due to body or soul?3 Evil (kakon) for Plato exists both in 
humanity and in nature as a whole, and covers physical as well as moral 
1 Pace Cherniss (I.e. 253 n. 34), P. does say at 509 b6-8 that the Form of Good is responsible 
for the existence and nature (?? ????? ?? ??? ??? ??????) of the objects of knowledge (sc. the 
other Forms). His citation of 517 c 3-5 against it ignores the fact that ??????? can (and in this 
context obviously does) mean reality rather than truth (p. 69 above). 
2 The identification of lack of goodness with its direct opposite, badness, at Rep. 353 c may be 
instructively compared with its heir, the Aristotelian conception of 'privation* or lack of form 
(???????) which though called 'what essentially is not* (Phys. 191 b 15), at the same time 'is in a 
way form* A93b 19) and 'works harm* A92315). In the first quotation 'is not* does not refer to 
existence but means 'essentially is not x': coldness being essentially (???* ????) not hot can never 
become hot, but a cold particular may. Aristotle maintained the distinction between 'opposites* 
and concrete things which 'possess* the opposites no less than P. when he wrote Pho. 103b-c. 
3 I do not think we should look for P.'s mature convictions about the relation between body 
and soul at Charm. 156c. There S., to make the purely Socratic point that 'care for the souF is 
more important than care for the body, quotes an imaginary Thracian sage, from whom he claims 
to have obtained a herb which will cure Charm.'s headache, as saying that ' all good and evil in 
the body and in the whole man spring from the soul*. P.'s touch here is light, and his point is 
simply to draw the familiar analogy between health and moral education. (Cf. vol. iv, 164.) The 
Charm, is a relatively early dialogue, and contains no trace of the developed doctrine of Forms. 
Nevertheless it is not inconsistent with the conclusion suggested here. Without soul, the body 
would be a lifeless lump, incapable of action of any kind. 
93 
Parmenides, Theaetetus, Sophist, Polidcus 
defects, in sentient beings such phenomena as disease and pain, in 
inanimate objects lack of the power to function well (cf. the pruning- 
knife), and in the cosmos as a whole irregular motions leading to  
confusion and disarray. On the cosmic scale it is considered mainly in the 
later dialogues. The Phaedo teaches that men may commit sins but only 
because the soul is corrupted and made to forget the Forms by its 
association with the body, which is itself evil F6b-e et ai.). In its 
discarnate purity, face to face with the Forms, soul is perfectly good. 
The effect of physical condition on moral disposition is graphically 
described in the Timaeus (86b ff.). The Republic develops the picture 
with the notion of internal conflict between three impulses in the 
incarnate soul, and in the Sophist B27e-28e) Plato posits two types of 
psychic evil, conflict and ignorance, which he compares to disease and 
deformity in the body. The latter call respectively for medicine 
and gymnastic, to which correspond, for the soul, punishment and 
instruction.1 
So much for moral evil in the human sphere. The whole cosmos, 
though designed by divine reason on the model of the Forms and  
therefore as good as it can be, has the faults inseparable from physical 
realization in space. (So the Timaeus.) The myth in the Polidcus 
B69 c ff.) explains that, though it is the best and most regular of all 
living things, and therefore endowed with the most nearly perfect of 
motions, circular revolution in the same place, yet because it has a body 
its motion cannot continue the same for ever. When therefore it has 
revolved for aeons in one direction under the hand of God, he  
withdraws his control and it reverses its movement, following its own 
innate impulse as a living creature. Two possibilities are explicitly 
denied: that God could move it in opposite directions is not meet 
(?????) nor can the alternation be due to ' two gods of opposed minds'. 
When under its own control, it gradually forgets2 its creator's teaching 
1 In the Laws (860 cff., pp. 376 ff. below) P. argues that the need for punishment does not 
invalidate the proposition that all wickedness is involuntary and due to ignorance. The seeming 
distinction between voluntary and involuntary wrongdoing must be explained in some other way 
(861 c-d). Even here he is careful to say that of the two kinds of badness one is called by the many 
wickedness while they call the other ignorance. 
2 ????? ???????????? 273 c. Cherniss (I.e. 27) observes that in general it is forgetfulness of the 
Forms which causes a soul to do evil. One may perhaps recall that in the myth of Rep. 10 the 
souls make their choice of lives (in which some do badly) before drinking the water of forgetful- 
94 
Theaetetus 
' owing to the bodily element in its composition' B73 b), and all sorts of 
evils spring up and threaten to destroy it, until God, to prevent this, 
takes control once more. Thus on the cosmic scale too Plato attributes 
evil and destructive influences to the body. 
It is in his last work, the Laws, that his position seems to have changed. 
His purpose in bk 10 is to combat the lack of moral standards resulting 
from a current form of materialism and atheism which saw the whole 
world as a product of chance. Nature is inanimate and purposeless, gods 
are human inventions, law and morality artificial and unstable, and the 
best course is to get all you can at others' expense, by force if necessary 
—the familiar farrago which he has so often attacked before. Here 
however his attack culminates in a foretaste of the astral theology 
developed, whether by Plato or another, in the Epinomis. All that  
concerns us now is that he starts from the affirmation of the Phaedrus that 
psyche, soul or life, as the only thing capable of spontaneous motion, is 
the ultimate cause of all motion or action everywhere. As such it is the 
cause of 'all contraries, good and evil, just and unjust, fair and foul' 
(896 d), and there must be at least two souls concerned, one working 
good, the other evil. (This is stated without argument, but follows from 
the fact that, as he will mention shortly, the souls of the astral gods are 
wholly good.) But the good and intelligent type is in supreme control, 
for the primary cosmic movements which govern all the rest, namely 
circular revolution, are the physical manifestation of intelligence 
(897c). 
There is here no Zoroastrian or Manichean dualism of God and 
Devil.1 Plato speaks of'kind of soul' (????? ????? 897b)  
interchangeably with 'sour, and thinks of the heavenly bodies as living creatures 
each with its own soul like earthly animals.2 The denial of two opposed 
gods in the Politicus is not contradicted. Whether or not Plato has 
renounced his previous view that evil is due to the body, or matter, is a 
difficult question which I cannot claim to have decided. Cherniss's 
ness. But (apart from the dangers of looking for allegory in every detail of a Platonic myth) the 
souls in question have all had a previous incarnation and are not fully purified. The effect of their 
former life still clouds their judgement. (Cf. vol. iv, 558f.) 
1 Cf. p. 365 below and Koster, Mythe de P. etc. 36 f. 
2 'Soul' is a generic or collective term as well as a singular one, and since Greek lacks an 
indefinite article P. was not obliged to mark the difference. 
95 
Parmenidesy Theaetetus, Sophist, Politicus 
solution is the outcome of deep thought and a comprehensive survey 
of the evidence. He concludes that evil for Plato is of two kinds, 
negative and positive. Negative evil is of course the necessary  
consequence of the falling-off of the physical world from the perfect 
reality of the Forms. Positive evil is caused directly by soul acting 
intentionally but in ignorance, and indirectly by the unintended 
effect of the good motions which it imparts, and which accidentally, 
by the necessity inherent in physical bodies, act on other physical 
phenomena. Why soul should ever lapse into error and ignorance of 
the Forms is a question to which Plato cannot be expected to give 
an answer.1 
I should be the last to claim that the 'problem of evil* can be solved 
in other than mythical language.2 But is it too ' unitarian' a view of 
Plato to suppose that, in the Laws as in the earlier works, soul only 
works evil when corrupted by the body? The Laws agrees with the 
Theaetetus (and the myth of the Phaedrus) that there can be no evil 
among the gods. The gods there are sun, moon and stars, and their 
souls are perfectly good (899b). Though soul is the cause of all good 
and evil, the statement that only good soul is in complete control, and 
the description of the souls of the astral deities, show that it can only 
work evil in the sublunary world; that is, the world of transient, 
physical phenomena, where the infection of the corporeal brings on 
forgetfulness and tempts it to act wrongly.3 When Plato says that at 
least two kinds of soul must be at work in the management of the 
ouranos, he adds at once that this includes earth and sea. It is in applying 
its psychical characteristics to the secondary motions of material  
substances—processes like growth and decay, mixture and separation, 
heating and cooling, qualities like hardness and softness—that soul may 
exhibit folly as well as wisdom and produce ill as well as good. It 
remains the moving force in everything, but the direction of the motion, 
1 See Cherniss's important article cited on p. 92 n. 2. It should be read in full. 
2 Cf. my essay on the soul in P., Entretiens Hardt vol. in, 141". 
3 So Tht. 17637-8, ????? ??? ?????. The 'visible gods' G7m. 41a, Epin. 985d) too have 
bodies, but of a substance not subject to destruction (they are everlasting, Tim, 40 b) or to any 
irregularity of motion. They move (P. believed) in perfect circles, the only motion that can 
continue indefinitely, and the analogy in the visible world to pure intelligence in the psychical; 
therefore they (that is, their souls) can be entirely rational. The Tim. explains all this at greater 
length. Cf. also Epin. 982c. 
96 
Theaetetus 
whether to good or bad ends, depends on how far it resists the  
corrupting influences of the body.1 
(iii) Are there Platonic Forms of evils? This much-debated question is 
even less susceptible of an answer than the last. The dialogues are not 
systematic treatises, and there are limits to the extent to which they can 
legitimately be synthesized. But two things can be said at once: first, 
the question was of no great interest to Plato; second, at no period did 
he allow a place for evil of any kind in the realm of the divine, which 
was the home of the eternal, changeless Forms. 
Plato disliked technical precision in the use of words (Tkt. 184 c) and 
even his key terms can be multivocal. In the unlikely event of a modern 
philosopher promulgating the Platonic doctrine, he would undoubtedly 
take pains, by using a different word or symbol, to distinguish the 
technical sense of Forms from that in which the non-philosopher 
naturally says ' There are many forms of evil.' Plato's thought may 
have been itself affected by the homonymity. At any rate he used the 
word eidos in both senses,2 and it is not always easy to know which is 
intended. For instance, at Rep. 402 b-c he says that an educated man 
must be able to recognize the eide of self-control, courage, liberality and 
high-mindedness and their kindred, and also their opposites, as they go 
about everywhere,3 and perceive both them and representations of them 
... believing them to belong to the same science. Although the Forms 
are prominent in later books, this need not refer to anything more than 
the 'dividing according to kinds' already attributed to Socrates, and 
several considerations suggest that the transcendent Forms are not in 
his mind here.4 (a) He is describing the primary education of the whole 
1 For different views see Solmsen, P.'s Theol. 141 f., Grube, P.'s Th. 146 k and the reff. they 
supply. That the bad type of soul is not (as G. thinks) confined to human souls should be obvious. 
Cf. also Dodds inJHS 1945, 21. 
2 Examples (among many) of ????? as sort or kind are: Phaedo 100b ??? ?????? ?? ?????; Rep. 434b 
(?? ??? ????????? ?????), 44°d-4la; 44Ic ?^?°5 used synonymously with ????? (see Adam on 
435 b); Laws 963 c ?????? ????. A good example of the ambiguity of the term is Phdr. 249 b, where 
???????? ???' ????? ????????? describes a purely logical process, yet to be capable of it a mind must 
have seen the Forms, and the phrase may mean either (as the continuation suggests) 'spoken 
of in generic terms' or 'called after a Form'. (Absence of ?? makes the former more probable.) 
3 ???????? ????????????, not quite the terms in which P. would describe the behaviour of his 
Forms. 
4 Many scholars, from Zeller (n.i, 560 n.) onwards, have assumed that they are, e.g. in recent 
times Grube, P.'s Th. 21, and Cornford in his translation. 
8 
97 
GHG 
ParmenideSy Theaetetus, Sophist, Politicus 
guardian class, if not, as many think, the whole citizen body. (Cf. vol. 
iv, 455-7.) Yet only a select few, after fifteen years of adult education, 
will advance to recognition of the Forms, (b) ' Representations' or 
'images' (???????) of eide might suggest instances of courage etc. in 
action, which with physical phenomena are so often called copies of 
Forms, but as Adam saw {ad loc.) Plato's subject here is education in 
poetry and art and the reference must be to narrative or graphic  
representations inspiring emulation of virtuous action or visible beauty. Such 
representations do not copy the Forms themselves, but only their 
earthly manifestations.1 (c) Even if the eide were Forms, Plato need not 
have been thinking of their opposites as Forms, rather than simply the 
lack of a Form.2 
Rep. 475 e-76a is more difficult. There ugliness, injustice and evil are 
called eide along with beauty, justice and goodness. It would of course 
make sense to translate eide as qualities or even concepts. Thus (to 
paraphrase): ' Beauty and ugliness are single concepts opposed to one 
another, and so are justice and injustice, good and evil. Every concept is 
single, yet they are associated with innumerable particular things and 
actions, as well as, in some cases, with each other. The philosopher is 
one who does not merely perceive the particulars but grasps the  
concepts. ' The rest of the book, however, leaves little doubt that eide here 
are Plato's independently existing Forms.3 On the other hand, in the 
ensuing comparison between the lover of sights and sounds and the 
philosopher the eide of ugliness and evil are simply ignored, though 
there are places (like 479 a) where mention of them would be quite 
appropriate.4 As to Tht. 176e, the expression 'patterns established in 
reality' (?? ?? ???? ??????) of godly happiness and ungodly misery 
seems to me to recall so strongly the language of Forms that Plato must 
1 This passage has been more fully discussed in its proper place, vol. iv, 545 ff. 
2 So Rist, Phoenix 967, 291. 
3 It is hard, as Rist (I.e. 291) has said, to see how a Form of Evil or Injustice by itself could 
exist at all, since the Form of the Good is cause of the existence of the objects of knowledge 
(Forms), so that in so far as they exist they must be good. (Cf. p. 93 above.) Yet it is equally hard 
to agree with him that we have here too nothing but a '"lack" of the Form of Good'. 
4 It is tempting to add that when beautiful particulars are said also to seem ugly, this is 
attributed to the fact that they are ' between being and not being', which might support the idea 
that evil and its kin are nothing positive but only a lack of goodness. But does this apply to one 
member of a pair like greatness and smallness, weight and lightness, cited as parallels to goodness 
and badness, justice and injustice? One hesitates, for fear of falling into S.'s 'pit of nonsense'. 
98 
Theaetetus 
at least be seeing both for the moment as absolute and changeless; but 
paradeigmata are not always heavenly like the Form of the State {Rep. 
596b),1 and Rist {I.e. 290f.) would confine 'reality' here to this world, 
on the ground that the man who models himself on the bad and godless 
pattern will be condemned to this world and never enter the divine one, 
which is 'a place pure of evil \ That, we must agree, is the 'place above 
the heavens' where the Phaedrus locates the Forms. 
Again at Rep. 445 c Plato seems to slip from one of the related but 
not identical uses of eidos to the other, when he says that the Form of 
arete is one, but those of evil are infinite in number. Virtue is a Form, 
and there are also Forms of the separate virtues. (In the Laws, 963 a ff., 
he discusses the old question of the Protagoras, how it can be right to 
call Virtue one and four at the same time.) But he would hardly posit a 
Form of something that has an infinity of forms or varieties. Forms are 
above all knowable, and the infinite is unknowable. 
There are two main reasons why it may have seemed logically  
necessary to Plato (when he remembered, and when he did remember he 
appears to have momentarily forgotten some of the most important 
aspects of his doctrine of Forms) to include evil among the Forms. 
{a) Forms, apart from their independent existence as steadfast goals of 
becoming, undoubtedly retained the functions of the universals, or 
common natures, which they originally (i.e. with Socrates) were. There 
is an eidos for every group of particulars which have the same name, 
because (as Socrates insisted) to use a single name is to assume a  
common nature {eidos) among the things it names, which in Platonic terms 
means that they share in the being of a single Form.2 {b) Secondly, Plato 
shared the general Greek tendency to see the world in terms of oppo- 
sites. This polarizing habit was common to the early Ionians, Heraclitus, 
Parmenides, Empedocles, Anaxagoras and the Pythagoreans, whose 
'table of opposites* (vol. 1, 245) may have influenced Plato directly; 
and it still prevailed in Aristotle, for whom the term ' opposites' was a 
frequent alternative to ' forms', since for him forms normally occurred 
1 The word is of course a common one. In Plato himself, S. as 'wisest of men' was held up as a 
?????????? by Apollo (Apol. 23 b), and he gave Meno a ?????????? of a definition {Meno 77 a); 
a ?????????? of bad and good oratorical technique appeared in the speeches of S. and Lysias in 
Phdr. B62 d), and so on. 
2 See Rep. 596 a and vol. iv, 550. 
99 
8-2 
Parmenides, Theaetetus, Sophist, Politicus 
in pairs of contraries like hot and cold. Moreover to know a form was to 
know its opposite.1 
To sum up, we are left in much uncertainty, because the dialogues 
never squarely face the question of the status of evil vis-a-vis the Forms. 
Hints that Forms of evil and ugliness exist occur only in passing and 
always in a context of opposites. Plato's attitude is that of his Socrates 
when asked whether there are Forms of mud, hair and dirt. To suppose 
there are seems absurd, yet he is sometimes tormented by the thought 
that what is true for one should be true for all (' Forms of every named 
group'). But at that point he runs away,2 and reserves his energies for 
what is worth while, the Forms of moral and aesthetic values, summa 
genera like unity, motion, similarity and their opposites, mathematical 
concepts and natural species—whatever in his view has a positive place 
in a system ideologically organized—reminding himself that Forms 
exist in a timeless, divine region to which evil has no access. These were 
quite enough to provide him with problems, as the Sophist will show.3 
Return to (i) 
Final attack on the theory of total flux A7^-83 c). Having allowed  
ourselves our own digression, we must return to the present argument. 
Protagoras failed the test of prediction, but it may yet be true that 
immediately present sense-experiences, and the beliefs or opinions 
(doxai) based on them A79 c 3-4), are infallible and constitute  
knowledge. This therefore, Socrates continues, they must examine further, 
not forgetting that there is a school of thought (the Eleatic) which holds, 
in diametrical opposition to the Heraclitean, that any form of movement 
or change is impossible. 
1 Cf. p. 90 ?. ? above. 
* ?????? ??????? (Parm. ?^??) is a very strong expression. 
3 Cherniss (Lc. 27, n. 34) lists a number of passages in which he sees Forms of positive vices. 
Some of these have been discussed here, others by Rist, Lc. 289-93. On Euthyphro 5 d I hope I 
have shown in my ch. on that dialogue that the doctrine of transcendent Forms had not yet 
taken shape in P.'s mind. Parm. i30C5~e4 does not mention Forms of bad things but only of 
trivia, 'things undignified and worthless' (though hair and clay are useful enough, and certainly 
do no harm). I do not see Platonic Forms at Soph. 251 a or Laws 964c. Bodily diseases—indeed, 
as we know, the body as such—can affect the soul G7m. 86bfT.) and impede its thinking (Phaedo 
66 c 1), but are not necessarily an evil (Rep. 496 b-c). In any case Forms of diseases are not 
mentioned at 77m. 87b-c, to which, with Phaedo 105 c, Cherniss refers us. (Tailpiece: Asclepiades 
reports Aristotle as saying in his Platonic Discussions that 'we (that is, the Platonists) say that 
there are no Forms of evils', Arist. Frr. ed. Ross p. 113.) 
IOO 
Theaetetus 
Those who assert that all is motion include in that term not only 
local motion but also alteration,1 for they will not concede that reality 
is static in any respect. Everything is always changing in every way (cf. 
p. 79). But then sensation is no more sensation than not sensation, so 
if knowledge is sensation, it is no more knowledge than not knowledge. 
If sensation changed its content but kept its character as sensation, it 
would be in at least one respect unchanging, and so transgress the 
neo-Heraclitean law of flux.2 In fact any answer to any question will be 
both correct and incorrect, and no existing language can express their 
thoughts. On the 'cleverer* theory of how sensation works it cannot be 
supposed to be knowledge. 
To the surprise of the others, Socrates, who himself introduced the 
Eleatics into the conversation, is unwilling to discuss their views A83 d- 
84b). His early meeting with Parmenides left him with such an  
impression of intellectual depth and nobility that he fears to misinterpret him. 
Besides, it could distract him from their main purpose, to discover the 
essence of knowledge by assisting Theaetetus to give birth to the 
thoughts with which he is in labour. 
Final disproof of the identification of sensation with knowledge: the role of 
thought (i84b-86e). A man is not a kind of Trojan horse in which the 
sense-organs lurk as individuals, the eye seeing, ear hearing and so 
forth. They all converge on one thing—call it the psyche or what you 
will—which employs them as instruments in making the man aware of 
perceptible objects. Each bodily organ conveys only its own kind of 
object—the eye colours, the ear sounds—yet we can think of the objects 
of several at once, e.g. of sound and colour that they exist, are different, 
are two and so on. Such concepts—being and non-being, similarity and 
difference, same and other (and, adds Theaetetus the mathematician, 
odd and even and numbers in general)—the psyche perceives not 
through any bodily organ but by itself. The same applies to aesthetic 
1 No English word covers the same ground as kinesis, translated 'motion'. It includes every 
sort of change as well as motion in space. Aristotle recognized four kinds: local motion (????), 
qualitative change (????????), change of size (?????? ??? q>0fais) and fourthly coming-to-be 
and perishing (?????? ??? ?????). 
2 i82d8-ei. The hypothetical objection stated and met by Cornford (PTK 98 f.) seems to 
depend on the concept of 'moments' of a discrete, Zenonian kind, which the 'flowing  
philosophers' would not admit. 
??? 
Parmenides, Theaetetus, Sophist, Politicus 
and moral qualities ('fair and foul, good and bad', 186a8). It is not by 
passively receiving what the senses convey, as any child or animal can, 
but by actively comparing and reasoning about them—which demands 
effort, education, maturity—that the psyche (mind) grasps reality and 
truth, without which no one can be said to know. Therefore sensation 
and knowledge cannot be the same thing. 
This account of the mind reaching knowledge of the truth {aletheia) 
by reflecting on sense-experience might suggest at first sight an  
empirical theory of knowledge, but that is far from Plato's thoughts. The 
language of the passage illustrates once again the use of alethes and on 
(with the nouns aletheia and ousia) as practically synonymous, to 
signify what exists or is real. (See p. 69 above.) * Sensation cannot 
reach truth because it cannot reach Being' (i86e4). 'Being' {ousia) is 
the key word in the argument.' The psyche will perceive through touch 
the hardness of one thing and the softness of another, but their being1 
—what they both are—their contrariety and the being of the contrariety 
. . . it essays to judge for itself A86b). The distinction between a 
sensible quality and its ousia was made in the Cratylus D23 c, p. 47 
above): sounds, colours and tactile qualities are always changing, and 
hence, in the language of the Republic,' hover between what is and what 
is not'. What fully exists is their essence or Form.2 Similarly a hard and 
a soft apple exhibit in this respect a contrariety, as the mind perceives 
by going beyond the immediate sensations and comparing them 
(????????? ??? ??????????? 186b 8), but the hard one may become 
soft and that instance of contrariety vanish, whereas Contrariety  
remains an unchanging reality knowable through further operations of 
the intellect alone. If there were no other evidence that Plato retained 
the full theory of Forms when he wrote the Theaetetus, the use of ousia 
in this passage, and especially the duality of' the contrariety and the 
1 I.e. the being of hardness and softness (McDowell 191). 'Existence' Cornford, but in this 
context ????? must bear its other sense of' true nature', and probably, as McDowell suggests, 
??? ????? means ' what they are' or ' that they are [hardness and softness]', ??( being equivalent 
to 'i.e.' 
2 On P.'s belief in 'degrees of reality' see vol. iv, 487 ff. On the view taken here, P. does not, 
in Owen's sense (SPM 324 with n. 1), 'ascribe ????? to objects of perception': he says (as in 
Crat.) that besides audible sounds there is the ????? of sound, and so on, and that ???(? is what 
the mind seeks without the aid of the senses (i86a-b). Bluck in JHS 1957 (bottom of p. 182) 
makes the same confusion. 
102 
Theaetetus 
ousia of the contrariety', would suffice.1 As in the Phaedo, the senses 
can start the mind on its way to knowledge of reality, but not only must 
there be a mind to go further 'on its own'; it can only do so because 
there are unchanging realities to be known. Ontology and epistemology 
remained inseparable for Plato, and the ontology of the Theaetetus is 
that of the Republic. 
B) Knowledge as true judgement (doxa) 
Convinced that knowledge is 'not to be sought at all in perception' 
A87a), Theaetetus suggests that it is true doxa, provisionally described 
as the activity of the psyche when it investigates things without the aid 
of the senses. More precisely (i89e-c)oa), thought is silent speech, a 
debate of the psyche with itself, and its final pronouncement is its doxa 
—opinion, belief or judgement.2 In discussing its relation to knowledge, 
Plato is resuming a theme already familiar from the Meno and 
Republic^ 
The main question is quickly, indeed cavalierly, disposed of B00 d- 
201 c) by an analogy from the law-courts illustrating precisely the same 
truth as the analogy of the road to Larissa in the Meno, namely that 
knowledge must be first-hand, something seen by the knower, not 
merely reported to him. By his choice of analogy Socrates also gets in 
yet another thrust at his betes-noires, the forensic orators, in substance 
a repetition of the Gorgias. In the short time allowed them, they cannot 
instruct the jurors about the facts of a case, but only persuade them to 
certain doxai about it.4 What they persuade them of may be the truth, 
but only an eye-witness could know that it was so. In the Meno (97a-b) 
the comparison is between a man who knows the way to a place be- 
1 It is of course perfectly possible to conceive the duality as simply that between the particular 
instance of contrariety observed in the case of the hard and soft apples and the concept of  
contrariety abstracted from that and other instances by the mind. That indeed is how it would appear 
to everyone today, and no doubt some will think that is how it must have appeared to P. I can 
only state my firm belief that for this to have happened to the author of the * middle dialogues', 
he would not only have had to undergo a credible change of mind; he would have become an 
entirely different person. Nor does ????? mean * concept'. If this is 'unitarianism', I am a 
unitarian. 
2 Gulley (PTK 87) prefers 'belief, McDowell (Tht. 193) 'judgement'. Both have their 
reasons. On retaining doxa see vol. iv, 262. 
3 See vol. iv, 256f., 261-4, 489-93. 
4 The distinction between instruction and persuasion, one giving belief, the other knowledge, 
was made at Gorg. 454cff., the point about shortness of time at 455 a. 
IO3 
Parmenides, Theaetetus, Sophist, Politicus 
cause he has been there before, and one who guesses it correctly. Both 
will get there, but one through knowledge, the other through true 
doxa; and the difference is of practical importance, for belief may be 
shaken, but knowledge never. 
I have called the examples both of the wayfarer and of the witness 
analogies,1 but some think them actual instances of knowledge as 
opposed to doxa, indicating a renunciation by Plato of his conviction 
that the object of knowledge cannot be any thing or event in the 
sensible world. Thus Stenzel wrote (PMD 71): 'The reason why . . . 
they [jurors] can have only right opinion is not that they have no 
knowledge of Justice itself (which, at an earlier time, would have been 
the reason given), but that they were not eye-witnesses of the crime. 
So that even in ethics the same entire change in Plato's view of  
knowledge is noticeable.' So too Hicken (Phron. 1958, 140) adduces the  
eyewitness as evidence that Plato has so far changed his position from the 
Republic as ' to bring the perceptible world within the range of  
knowledge'. For Robinson (Essays 41) this apparent denial of the Republic's 
view is either a slip or an unnoticed implication on Plato's part. For 
A. Rorty (Phron. 1972, 228) the eye-witness example suggests that the 
objects of knowledge and doxa may be the same.2 This last suggestion 
is not quite true. The witness's knowledge is of the crime itself, the 
juryman's doxa is of the witness's mimesis of that crime in words, a 
good analogy for the difference between the knowledge of a Form and 
of its mimesis in a sensible object or action. 
There is a way of allowing that Plato spoke of knowledge (episteme) 
of the sensible world without implying any volte face on his part, namely 
by assuming that he sometimes used the word more generally, or in two 
senses. Thus Runciman (PLE 38): 'Eye-witness knowledge is not, of 
course, an instance of that highest and truest knowledge which... Plato 
distinguishes from phenomenal knowledge as in the Parmenides, 
Phaedrus and Philebus. But it is a perfectly good example of the  
knowledge which can be acquired . . . within the ontological frontiers of the 
empirical world.' And Rist (Phoenix 1967, 284) cites the road to Larissa 
1 For the former cf. vol. iv, 240 n. 3. 
a Similarly Stoelzel, 11 n. 2: Right doxa is distinguished from knowledge ' nicht so sehr durch 
den Inhalt als vielmehr durch die Art der Entstehung'. 
IO4 
Theaetetus 
and the witnessed crime as in Plato's eyes legitimate examples of 
episteme. The strongest evidence for this is Philebus 6id-e: 
We agreed that one pleasure is more truly pleasure than another, and one art 
more exact than another. And knowledge differed from knowledge, one 
directed to the things that come to be and pass away, the other to those that 
do neither, but exist for ever, constant and unchanging. Examining them 
from the point of view of truth, we concluded that the latter was truer than 
the former. 
Plato, then, in the later dialogues occasionally spoke of two kinds of 
episteme corresponding to what he also called episteme and doxa\ one 
directed to unchanging reality, the other to unstable phenomena, and 
one * truer' than the other. Here is no 'entire change of view' or  
abandonment of the Forms. Similarly the whole tenor of the Theaetetus, the 
manner in which the claims of sensation are dismissed, and the aporetic 
ending, point to the abiding necessity of the Forms if'true' knowledge 
is to be attainable. It makes no essential difference whether we call the 
state of mind of the eye-witness and the experienced traveller analogous 
to knowledge or knowledge of an inferior grade,1 but I believe Plato's 
meaning has been best expressed by Bluck {Mind 1963, 260): 
Knowledge is to true belief as the state of mind of an eye-witness is to the 
state of mind of a juryman who is won over by persuasion. We are inevitably 
reminded of the road-to-Larisa illustration in the Meno . . . Both analogies 
suggest that some sort of personal acquaintance is the mark of knowledge;2 
and as applied to a priori knowledge, the Meno illustration certainly meant 
that knowledge involved yvcoais [cognition] of ?? ???? [the things beyond]. 
It is natural to suppose that the Theaetetus analogy, as applied to such things 
as existence and likeness, ought to mean the same. Furthermore, the Timaeus 
tells us that if knowledge and true belief are different, then there are Forms, 
whereas if they are not different, sensible objects must constitute reality 
E id). Even if the Timaeus preceded the Theaetetus, it would seem natural, 
in the absence of clear evidence to the contrary, to take the present distinction 
as implying that knowledge is of Forms. 
1 It is extremely important with Plato to distinguish confusion or change of mind from his 
simple dislike of technical or pedantic language. Just as he occasionally uses ???????? in the late 
dialogues for what he elsewhere calls ????, so he uses ???? for what, were a philosophical point 
involved, he would call ?????????, e.g. at i88e7ff. (For this, see p. 232 below.) 
2 On the language of direct vision as applied to the apprehension of Forms cf. vol. iv, 252, 
392, 507, 511. The soul has been an eye-witness of them. 
105 
Parmenides, Theaetetus, Sophist, Politicus 
When at Tim. 51 d3 ff. (wrote Ross, PTI i24f.) Plato says that the 
existence of the Forms depends on the difference between knowledge 
and true judgement, he is relying on the argument here in the Theaete- 
tus, 'in which he claims to establish just this difference'. This shows a 
better understanding of Plato than Stenzel's. 
But is false judgement possible? (i87d-20od).1 The notion of true  
judgement implies a contrast with false, and before dismissing its claim 
Socrates initiates a long, complicated and fruitless digression2 on the 
nature and possibility of false judgement, only to conclude that it was 
wrong3 to consider it before settling what knowledge itself is. 
Earlier Plato had been content, like anyone else, to accept the existence 
of false beliefs as an obvious difference between knowledge and doxa. 
(Cf. Gorg. 454d.) Now, with his shift of interest towards logical and 
epistemological problems, he feels that the paradoxes of men like 
Protagoras, Antisthenes and Cratylus4 are not to be so lightly brushed 
aside. The solution is only reached in the Sophist (pp. 154-6 below), 
(i) False judgement as mistaking one thing for another A87 c—88c). 
Here Socrates deliberately adopts the Sophistic starting-point by  
ignoring 'for the present' learning and forgetting as intermediates between 
knowledge and ignorance;5 and the question itself treats knowledge as 
confined to acquaintance with a person or thing. On these premises false 
judgement is quickly rejected on the ground that one cannot mistake 
1 W. Detel has devoted a monograph to the subject of false judgement, or false statement, in 
Tht. and Soph., which begins with a useful survey of over a score of recent interpretations and has 
a full bibliography (Platons Beschreibung des falschen Settles im Theatet und Sophistes, Hypomne- 
mata 36, Gdttingen 1972). His own thesis is that, contrary to the usual view, Plato did not see 
himself as faced simply with the same difficulties as the Sophists and others who denied the 
possibility of false judgement, but rather with difficulties arising solely from his own theory of 
Forms and the way in which, in consequence of that theory, he himself used the verb 'to be'. 
Detel posits a sharp distinction in Plato's mind between the first three accounts of false judgement 
and the last two (the similes of wax tablet and aviary). The last two dispose successfully of the 
others, which are those of the Sophists and other predecessors, but prove unsatisfactory for  
different reasons. Their problems, arising out of the theory of Forms, are solved in the Sophist. 
2 Not such a digression from the main question of the nature of knowledge (McDowell 194). 
3 B00d) An error of judgement? But this second-order difficulty, or virtual petitio pr'tncipii, 
is (rightly enough) ignored. 
4 For these see vol. m, 182, 207, 210f. 
5 A88a) The tactic on which the fighting brothers relied in Euthyd. See 275 d~77c, 277e-78b. 
Learning and forgetting are restored to their place in introducing the simile of the mind as a wax 
tablet A91c). 
IO6 
Theaetetus 
one person for another whether one knows both or neither or one but 
not the other. 
(ii) False judgement as 'thinking what is not' (i88c-8c)b). This  
suggestion takes us back behind the Sophists to their Eleatic original.1 It 
fails because it is impossible to think what is not, as Parmenides had 
said (frr. 2.7-8, 8.8-9). Plato however goes beyond Parmenides when 
he adds to * think what is not' the words * whether about any existing 
thing or absolutely' (lit. 'by itself). Parmenides could not speak of 
non-being in relation to ' any of the existing things' (??? ????? ???) 
because only a single Being existed,' one, continuous ... by itself (fr. 
8.6, 29). However, this is not necessarily (as Bondeson suggestedJ an 
anticipation of the distinction to be drawn in the Sophist between the 
two senses of'what is not', the existential and the merely differentiating 
—' does not exist' and' is not ?' (sc. what it was wrongly thought to be). 
The present distinction is more probably between thinking 'it is not' 
within the Parmenidean scheme of one unique Being (that 'wholly 
undiscoverable path', fr. 2.5f.) and thinking of one of the many 
commonly accepted existing things that it 'is not' in the same sense (i.e. 
does not exist).3 Anyway, by explicitly refusing to question the Eleatic 
thesis A83 d 10 ff.), Plato has made it clear that this indispensable means 
to an understanding of 'what is not' is to be reserved for the Sophist, 
where 'father Parmenides' will be cross-examined and the conclusion 
enforced that 'what is not in some respect is' B41 d). 
The argument here rests on a simple analogy between sense- 
perception and judgement: if one sees, hears or feels something, there 
must be something which one sees, feels or hears. Similarly if one judges 
something there must be something that one judges. One cannot  
therefore judge 'what is not', for one's judgement would then have no 
object, one would judge nothing, and so not be making a judgement at 
all.4 This therefore cannot be the explanation of false judgement. 
1 They used it too, of course (Euthyd. 283 a ff.), but S. does not draw the Sophists' conclusion 
(i89b4). 
a Phron. 1969, 1171". B.'s is a most interesting and suggestive article, though one would have 
liked to know what he made of Bluck's '"Knowledge by Acquaintance" in P.'s Tht.' in Mind 
1963. 
3 Cf. Stoelzel's rendering (p. 86): ' sei es als das Sein eines Gegenstandes oder als das Sein an 
sich'. 
4 Cf. the similar argument about speech at Soph. 237d-e. 
IO7 
ParmenideSy Theaetetus^ Sophist^ Politicus 
The persistence of the problem discussed here by Plato is brought 
home by R. M. Gale's article on propositions and judgements in the 
Encyclopaedia of Philosophy, VI, 494-505, which is largely concerned 
with theories about whether what a person thinks of must have some 
independent reality, and what is the object of a false judgement. Thus 
Moore wrote (see ib. p. 496): * In order that a relation may hold between 
two things, both the two things must certainly be; and how then is it 
possible for anyone to believe in a thing which simply has no being? ... 
I confess I do not see any clear solution of the difficulty.' According to 
Gale, the false analogy in the Theaetetus (he quotes 18933^2) has 
haunted most modern theories of judgements and propositions except 
the behaviouristic. He distinguishes two classes of verbs: propositional 
(such as * judge', * think', 'believe') and what he calls cognitive (such as 
'know' and the verbs of sensation, 'see', 'hear', 'feel', 'taste', 'smell').1 
Cognitive acts require objects, but propositional acts do not. If one sees 
a cat on a mat, or knows that it is there, there must be a cat, but if one 
only thinks there is, there may be none; which means, on this view, that 
there may be no object of a propositional act. 
(iii) False judgement as 'other-judgement'2 (i8c)b-9oe). Perhaps false 
judgement occurs ' when someone exchanges one real thing for another 
in his mind, and says it is that other. In this way he will always think 
what is, but one thing instead of another, and since he misses his aim 
can rightly be said to judge falsely.' This might be thought similar 
enough to (i) to be ruled out by the same argument^ but on the  
contrary Theaetetus hails it enthusiastically, and gives an entirely different 
kind of example, namely qualities: when someone 'judges "ugly"  
instead of "beautiful"', then his judgement is 'truly false'.4 Graciously 
waiving the oxymoron, Socrates does not interpret this in the obvious 
1 He does not mean that the objects of cognitive verbs cannot be propositional in form. 
2 P. apparently coined the word ????????? for 'misjudgement' on the analogy of ??????????, 
used by Hdt. A.85) of failure to recognize a person; perhaps also with the Homeric ?????????? 
in mind; its secondary meaning—to be knocked senseless—would appeal to his sense of humour. 
3 Ackrill has tried to show that the two arguments are not identical (Monist 1966, 388 ff.). 
His second point of difference at least goes against Cornford and R. Robinson. See the latter's 
Essays, 64 f. 
4 Both S. and Tht. take for granted the objective character of aesthetic and moral values as 
things about which one may be simply mistaken, on a par with odd and even, two and one. 
This I take to be the ordinary commonsense view, unrelated to the theory of Forms, that 'there 
is such a thing as justice'. Cf. vol. iv, 115 f., 223. 
IO8 
Theaetetus 
sense of judging a beautiful (fine) individual or action to be ugly 
(shameful). Instead, having established that judgement is the final stage 
of a mind's converse with itself (p. 103 above), he asks if anyone, sane 
or mad, has ever said to himself that beauty is ugliness, odd numbers are 
even, ox must be horse or two one. * Never', replies the bemused 
Theaetetus. 
There is surely some sharp practice here, exploiting ambiguities 
which cannot easily be reproduced in English. 'The beautiful' (?? 
?????), as we well know, may mean either what is beautiful or the 
quality of beauty (whether in the commonly accepted sense or as a 
Platonic Form). Theaetetus clearly had the former in mind, but one 
must assume that Socrates intends his examples to stand, primarily 
at least, for concepts or universals, partly because he speaks of'the ox' 
also with definite article (as when one says ' the ox is a patient animal'), 
but mainly because it is obvious that a man might mistake an ox for a 
horse in the dark or a schoolboy mistake 29 + 38 for 66, an odd number 
for an even. On Socrates's present interpretation, to make a false  
judgement as now defined one must consciously entertain the nonsensical 
statement 'An odd number is even' or oddness is evenness. The 
Theaetetus, though its lateness is scarcely in doubt, resembles the early 
dialogues in being deliberately aporetic and in consequence Socratic, in 
the sense that Socrates speaks in character and uses his own teasing 
arguments to avoid reaching a positive conclusion which nevertheless 
the reader can divine.1 His object, as with young Charmides, is not to 
teach but, by his art of mental midwifery, to elicit and test his  
interlocutor's own ideas. The gain lies, not in finding the right answer, but 
in purging the mind from error and the false conceit of knowledge 
B10c). To attain this end he is not above misrepresenting a young 
man's meaning, as in his shocking distortion of' to do one's own' in the 
Charmides A61 b; see vol. iv, 267). He does not shirk the real question 
in the end. Having enjoyed his little mystification, he pursues the  
problem of 'other-judging' more seriously in the similes of the wax tablet 
and aviary, where mistaken identity and arithmetical errors are taken 
into account and the sophistic exclusion of memory and forgetfulness 
is abandoned. At that later stage Plato allows Theaetetus to distinguish 
1 The most striking case of this is the search for courage in Laches (vol. iv, 132Q. 
IO9 
Parmenides, Theaetetus, Sophist, Politicus 
carefully the two cases here confused: * Seeing or touching n objects', 
he says, 'a man might think them 12, but he will never make that  
judgement about the 11 and 12 which he has in his mind' A95c). 
(iv) False judgement as the misfitting of a perception to a memory: the 
mind as a wax tablet A91 a~96c). One may certainly see a stranger at a 
distance and mistake him for Socrates whom one knows. This was 
impossible under the Sophistic limitation that one must either know or 
not know, and cannot both know and not know the same thing. That 
is now removed, having served the purpose of showing up the  
inadequacy of the Eleatic-Sophistic epistemology, and memory and  
forgetting are reinstated. Imagine the mind as a wax block, on which we stamp 
what we perceive or conceive,1 like the devices on seal-rings. So long as 
these memory-impressions do not wear out we know what we have 
perceived or conceived. Socrates lists exhaustively the cases in which, 
on this supposition, false judgement is impossible.2 The upshot is that 
both a present perception and a memory-imprint or concept must be 
involved, since false judgement consists in matching the perception to 
the wrong imprint. Seeing at a distance two men of similar height and 
build, both of whom one knows, one may, in fitting the perceptions to 
the memory-imprints, transpose them, like putting shoes on the wrong 
feet; or if one sees a stranger, wrongly match the sight to the imprint of 
an acquaintance. Socrates concludes by fancifully attributing disparity 
of intellectual gifts to variations of quality in the wax. Minds too soft 
learn easily but forget quickly, and hard wax, taking shallow  
impressions, also causes poor memory. Blurred imprints may result from  
softness, from adulteration, or from overlapping in a * small mind'. The 
best minds have large and deep tablets of smooth, well-kneaded wax, 
taking clear, lasting and well separated impressions. Such minds learn 
quickly, are retentive, and make true judgements, for they quickly 
assign the data of perception to the appropriate memory-imprints. 
Having obtained Theaetetus's enthusiastic agreement to his carefully 
? ???????, ???????. 'Conceiving is evidently intended to be a relation like perceiving, but 
with objects which are abstract, e.g. perhaps numbers' (McDowell 215). This seems to fit the 
context, though at Pho. 73 c ?????? as contrasted with the direct object of sight or hearing is the 
mental image of a person conjured up by the sight of one of his possessions. 
a They are fully tabulated in Stoelzel (97) and McDowell B15). Cornford gives a clear 
summary on p. 122 of PTK. 
IIO 
Theaetetus 
built up thesis that false judgement only occurs at the conjunction of 
perceptions with thoughts,1 Socrates proceeds to dismiss it as  
inadequate. True, we cannot judge that man is horse when perceiving neither, 
but we can confuse two unperceived concepts, e.g. numbers. To 
Theaetetus's sensible observation that one could mistake 11 objects for 
12 but not the one number 'in his mind' for the other, Socrates replies 
that one can wrongly think 7+5 (the numbers themselves, not 7 and 5 
objects) =11, and since 7+5 = 12, this amounts to thinking 12 is 11, 
and entails ' knowing what one does not know' in the forbidden sense. 
The substitution of G+ 5) for 12 is unfair. It is rather a truncated 
question than something known; at least it makes sense to ask 'how 
many are 7+ 5?' but not 'how many are i2?\2 Nevertheless it remains 
true that arithmetical errors are possible and are of a kind not allowed 
for on the wax-tablet model. 
(v) Knowledge potential and actual: the aviary (i97b-20oc). To 
escape this difficulty, Plato adumbrates what Aristotle has taught us to 
call the distinction between potential and actual.3 A man may possess 
knowledge in the sense that, having learned it, he has it stored in his 
mind, but not 'have it about him'. (He knows the names of his friends 
Taylor and Weaver, but with something else on his mind may refer to 
one by the other's name.) This suggests another metaphor. The mind 
is an aviary, full of birds of all sorts.4 The owner possesses them all, i.e. 
has a certain power over them: he can enter and catch one whenever he 
pleases, and will then actually have it. The birds are things known (lit. 
' knowledges'), to stock the aviary is to learn, and to catch a particular bird 
in the hand is to recall a thing once learned and so known in a potential 
1 ???????, as S. at 295c! 1-2 calls what he continues to call memory-imprints A9633). 'At this 
stage . . . memory is made to do the work of abstraction' (Campbell on 19633; cf. McDowell 
2 ' How many beans make 5 ?' W3S a question with which my grandfather's generation delighted 
to puzzle small children. 
3 Aristotle too uses knowledge as an example of the distinction, which can be threefold, e.g. 
at De an. 417b21: First, a man is potentially knowledgeable simply as being a member of the 
human race, which is capable of acquiring knowledge. Secondly, a literate man has a knowledge 
of letters, still in a potential sense, meaning that he can read or write whenever he wishes. Finally 
his knowledge is actualized when he is actively exercising these skills. (Cf. also Phys. 25 5 b 2, 
De an. 412a 10, Tht. 198c) These correspond, in terms of Plato's metaphor, to having an empty 
aviary, stocking it, and holding a particular bird in the hand. 
4 Nothing certain can be made of the addition that some are flying in flocks, others in small 
groups, and others singly. See McDowell 220f. or Cornford 132 n. 2. It is not referred to again. 
Ill 
Parmenides, Theaetetus, Sophist, Politicus 
sense. Thus one can know (potentially) what one does not know 
(actually), and here lies the possibility of false judgement: a man may enter 
his aviary meaning to catch a wood-pigeon, but lay hands on another 
variety.1 One who has learned the numbers 'knows' (is acquainted 
with) both ii and 12. If, when asked what is 7+ 5, he replies in good 
faith, ' 11', he has hunted in his memory for 12 but caught instead 11. 
The distinction between 'possessing' and 'having' knowledge,  
between having learned one's letters and being engaged in reading, is a 
genuine advance, and resolves the unreal dilemma about knowing 
and not knowing the same thing. Yet the aviary model, with its 
'knowledges' or 'pieces of knowledge' flying about like birds, is not a 
happy one, and is soon impugned by its author. Does it, he asks, make 
sense to say that, when it is a piece of knowledge that one has caught and 
is actually examining, one could fail to recognize it for what it is? In 
desperation Theaetetus suggests that 'pieces of ignorance' ('unknow- 
ings') may be flying about among the knowledges, and be caught by 
mistake. Then false judgement would consist in mistaking one thing 
for another, which is precisely what they have been trying, and failing, 
to explain from the beginning.2 It is hardly worth while to go into the 
question what the 'pieces of ignorance' ('misapprehensions', Ryle) 
might be. Indeed the aviary as a whole gives some support to Aristotle's 
dictum: 'What is expressed in metaphor is always obscure.'3 
It is often pointed out4 that one difficulty in accounting for false 
belief or judgement is Plato's assimilation of belief and knowledge to 
seeing or touching. This is basic to both his and Aristotle's epistemo- 
logy, for different reasons. Both thought of knowledge as acquired by a 
process resembling sensation in its directness, Plato because it consisted 
1 ? ????? ('ringdove' LSJ) for a ????????? ('common pigeon or dove'). A dove instead of 
a pigeon, say Cornford and McDowell, but see the Shorter O.D.: a pigeon is ' a dove, either wild 
or domesticated'. The choice of varieties so closely related makes the mistake sound easy and 
natural. 
2 See (i) and (ii) above. At 200 b there is an elaborate reference back to i88b-c. 
3 Top. 139b34. Sometimes, e.g. in trying to describe the nature of the soul (Phdr. 246a), or 
the relation between eternal Forms and their sensible instances, it may be the only resource 
available; but as an explanation of false judgement it is less appropriate. Two suggestions for the 
meaning of ??????????????? (the first is Cornford's) are mentioned by McDowell (p. 225), and 
Ackrill (Monist 1966, 400) tries to make sense of Plato's metaphor by yet another, of coloured 
and labelled cards. 
4 Recent examples are Sprute in Phron. 1968, 59, and Bondeson, Phron. 1969, 118. Cf. the 
phrase ?????? ??? ???????? ??????? at i86d. 
112 
Theaetetus 
in a sudden mental vision of a Form ensuing on the philosopher's 
reasoning about the objects of experience (p. 105 n. 2 above), and 
Aristotle because, after his abandonment of the transcendent Forms, 
the philosopher's grasp of immanent form or essence depended  
ultimately on the ability to make the first inductive (and rationally  
unjustifiable) leap from individual sensations to the lowest universal. This 
intuitive power he called nous, the highest of human faculties, yet at the 
same time the closest to sensation, or even identified with it.1 I mention 
all this now, in spite of the impossibility of discussing it fully, to 
emphasize that very much more is involved in Plato's arguments here 
than a mere vulgar error of confusing 'knowledge that' with  
'knowledge by acquaintance'. 
Two further points before we leave the aviary. First, it offers, like 
the wax tablet, an empiricist, tabula rasa view of knowledge, leaving no 
room for anamnesis of the Forms: we start with our aviaries empty 
A97c). Secondly, Socrates describes teaching as 'handing over'  
knowledge and learning as receiving it A98b). Together these suffice to show 
that for maieutic purposes Plato does not feel bound to express his real 
opinions. 
Socrates attributes their failure to explain false judgement to the error 
of attempting to do so before settling the question of the nature of 
knowledge itself.2 Returning then to the main question, they reject its 
identity with true judgement for the reason we have already seen (p. 103 
above). The rejection was in any case a foregone conclusion, for the 
distinction between knowledge and true judgement or belief was vital 
to Plato's thought from the Meno right through Republic and Timaeus 
to Laws F32 c). 
1 Metaph. 103635-8:' There is no definition of individuals, but they are cognized by sensation 
or intuition, and when we are not actually perceiving them it is not clear whether they exist or not. 
But they are always spoken of and known by the universal logos.* Cf. EN 1143b4-5: universale 
are made up of individuals. Of these one must have sensation, 'and that is nous*. Although the 
immediate object of sensation is the individual, sensation puts us in direct touch with the  
universal: seeing Callias we get our first awareness of man {An. Post. 100a 16-18). To intuit something 
is to touch it (???????? ??? ????, Metaph. 1072b21). 
2 200 c-d. Cf. i96dff. on the ' shamelessness' of attempting to say what knowledge is like 
before knowing what it is in itself. This methodical error takes us back to the Meno G1b, 
? ?? ?? ???? ?? ?????, ??? ?? ?????? ?? ?? ???????;, also Prot. 361 c). 
9 
113 
GHG 
Parmenides, Theaetetus, Sophist, Politicus 
C) Knowledge as true judgement with a logos B0ic-iob) 
Theaetetus now * remembers having heard' that knowledge is true judge- 
men or belief accompanied by a logos, with the corollary that only that 
can be known of which a logos is possible. That true doxa is converted 
into knowledge when one can 'give an account (logos)9 of it is Plato's 
own view as expressed at Symp. 202a; and Meno 98 a suggests that this 
consisted in a 'reasoning out (logismos) of the cause', equated with  
recollection of pre-natal knowledge. Here that thesis is put to a searching 
test, and found wanting whichever of three meanings one gives to logos. 
The theory dreamed by Socrates B01 d-2o6b). Theaetetus remembers no 
further details, so Socrates obligingly supplies them by relating a theory 
which he once 'seemed to hear',1 according to which the world  
(Ourselves and everything else' 201 e) is composed of complexes and their 
elements. Complexes 'have a logos\ namely a statement of their  
elements: elements have none, but can only be named. Speech reflects 
reality, and logoi are complexes of names corresponding to the complex 
objects which they define. Of an element it is not even legitimate to say 
'it is' or 'it is not', for to add being, non-being or any other predicate 
to it is to make it complex. Elements cannot be accounted for2 or known, 
but are perceptible. Complexes can be known because about them one 
can both have a true belief and give a logos. 
The theory is clearly not Plato's.3 It has a Sophistic ring,4 and most 
closely resembles one which Aristotle ascribes in his own terms to 
Antisthenes and his followers. At Metaph. 1043 b 28 he says that accord- 
1 I.e. in a dream: he will offer ???? ???? ????????$. Hence what follows is usually called 
'S.'s dream', and its significance sought by comparison with other metaphorical references to 
dreams in P. (See Burnyeat in Phron. 1970, 103 and A. Rorty, Phron. 1972, 229 k) The phrase 
sounds to me proverbial, meaning something like ' My story is as good as yours' (not the same as 
?? ???? ???? ?????? ???? at Rep. 563d). 
For a comparison of the theory with the elementary propositions of Wittgenstein (who 
discusses the Tht. theory in Phil. Inv. p. 21, §§46ff.) see McDowell 233 f. 
2 There is a play on dxAoyov, which in normal use meant 'irrational' or 'unintelligible', but in 
form is simply 'without logos*. 
3 Pace Hicken and Burnyeat, who have argued elaborately against Antisthenian authorship 
in Phron. 1958 and 1970. McDowell also thinks that Plato may have originated it (pp. 234, 237). 
4 Cf. the summary of fifth-cent, views in vol. m, 218 f. Pp. 209-18 give some background to 
the present discussion. The prohibition of attaching any predicate, pronoun or epithet B02a) to 
something that one perceives is typical of sophistry. 
114 
Theaetetus 
ing to them you cannot define what a thing is, but only say what it is 
like: * there is a kind of substance of which definition (????) or logos is 
possible, namely the composite, whether sensible or intelligible, but 
this is not true of its primary elements, for a defining logos predicates 
one thing of another \ His Greek commentator illustrates their difficulty. 
4Man' is a name. We may say he is a rational mortal animal, but this 
again is only a string of names. Even if * animal' can be divided into a 
further plurality of names, we shall finally come to a simple, elemental 
entity which cannot be so divided. This will be indefinable, and we 
cannot claim to have defined, or explained the essence of, something 
simply by describing it as composed of elements which are themselves 
indefinable.1 
Theaetetus agrees that this is the argument he had in mind, and they 
proceed to examine it. True belief and a logos are at least necessary  
conditions of knowledge, but as usual One thing' bothers Socrates. The 
unnamed thinker used the example of letters and syllables: the logos of 
the syllable 'so', which says what it is B0338), is *s plus o9; but one 
cannot give a similar logos of s or 0, naming their elements.2 They are 
mere noises, which can only be heard and named.3 But, asks Socrates 
1 See Antisth. fr. 44 b Caizzi. Hicken (/.c. 138) makes a point of the inconsistency between 
saying all definition is impossible and denying it only to simples. But since Aristotle himself with 
his ???? ??? appears to attribute both to Antisthenes, unless one regards the ???? clause 1043 b 28 ff. 
as no longer referring to him (which she does but I confess I find unnatural), I prefer to see in 
this addition an understandably careless expression such as Aristotle is often guilty of. The 
commentator's account (which she does not mention) is certainly that Antisthenes denied all 
definition. 
2 The word for letters and elements is the same, ????????. (?. is said to have been the first to 
use it in the latter sense, Eudemus ap. Simpl. Phys. p. 7.13 Diels.) More strictly ???????? are the 
elementary sounds of which letters like sigma are the symbols. So Arist. Metaph. 1000a 1-4, and 
in the Poetics A456b22) he defines them (if ch. 20 is genuine; see Bywater ad loc.) as 'atomic 
sounds' (????? ??????????). This they evidently are here, though neither P. nor Arist. always 
maintains the distinction. 
3 Burnyeat (Phron. 1970,119) says that to be unanalysable is not the same as to be indescribable, 
and in fact Tht. describes s as *a sort of hissing noise' B03b), which a pertinacious critic might 
even call an analytical definition by genus and species. McDowell B41) claims that a particular 
instance used in explaining the theory need not be an instance of ultimate, abstract elements as 
envisaged by the theory itself. Yet its author did use letters as paradigms (i.e. * instances of the 
sort of elements and complexes with which the theory is concerned', idem p. 240), and as Tht. 
says of letters (that is, of ????????, 203b2), 'How can one state the elements of an element?' 
It must be admitted that letters (or rather the elementary sounds of which they are the symbols) 
are treated here as primary elements B01 ci), which can only be perceived and named B02b). 
Incidentally the theory under discussion precludes what S. himself gave as a specimen definition 
at 147 c. 
 
9-2 
ParmenideS) Theaetetus, Sophist, Politicus 
(like his one-time follower Antisthenes), how can a complex of  
indefinable and unknowable elements be itself definable and knowable? Not 
if it is simply the sum of its elements, nor yet if their collocation has 
brought into being a new, unitary form,1 for that in turn will be an 
indefinable simple. In Hippias Major C01 b-c) Plato disproved the idea 
that a group and its separate members must have the same properties, 
by the example of numerical properties: neither a nor b separately is 
two, both together are two. So he knows that if neither s nor 0 can be 
known, there is no logical necessity that the two should be unknowable 
in combination. Here he refines on that argument. There would be such 
necessity if the complex were nothing more than the sum of its  
constituents B05 b). Only by assuming that the elements have fused into a 
new, single 'form' or * whole' can the conclusion be avoided, and the 
singleness of that form raises in this context the further difficulty that it 
too will be unknowable B05 e). 
The weakness of the theory is also shown up empirically, by the  
experience of learning. The elements—letters, notes in music etc.—are 
the basis of our knowledge of their complexes. In general, 'elements are 
more clearly known than their compounds, and more effective than the 
compounds towards a complete grasp of every subject', so that it is 
absurd to say that a compound is knowable and an element unknowable 
B06b). 
The 'dream-theory' interweaves what might appear to be two 
distinct, if related, questions, the logical one of how we can have 
scientific knowledge of individual members of a definable species and 
another of the relationship, from the point of view of knowability, 
between individual persons or things and certain postulated2 elements 
1 ??? ??? ???? 203 c, lv ????? 20364, ????????? ?? ??? ????????? 205 b. The language is that 
used of Platonic Forms in Phaedo (????????? 78 d, 80b, 83 ?), but here P. is speaking of elements 
in the phenomenal world, what the Phaedo G9a) calls 'things that one can touch and see'. The 
simplicity of an invisible, incorporeal form has up to now been the guarantee that it is not only 
immutable but knowable. Cf. p. 120 n. 3 below. Stenzel (JPMD 73) however thought ???? and 
????? here were Platonic Forms. Contrast the doubts expressed by Wedberg, PPM 143 n. 8. 
The argument of 2043-2053 involves persusding Tht. against his will—and fallaciously— 
that a collection of parts is the same as a whole—as if, as McDowell says (p. 145), having all the 
parts of a car were the same as having a car. At least their position makes a difference (Arist. 
Metaph. 102431). 'So' is not the same syltable 3s 'os\ 
2 In 3 commentsry on this pssssge (see p. 114 ?. ? sbove) Wittgenstein hss drawn 3ttention to 
the difficulty (of which P. of course had no suspicion) of pinning down a single use of the terms 
'simple' 3nd 'composite'. 
116 
Theaetetus 
of which they are composed. Simples or elements are not the same as 
individuals (Socrates and the dog Tray are highly complex), and it is 
difficult to know exactly what they are. One might suppose them to be, 
as in Antisthenes, the logical constituents of a definition, like 'rational' 
and ' animal' in the definition of the species man. Yet Plato calls them 
perceptible to the senses (aisthetd)? which may remind us of (though 
different from) Aristotle's view (referred to above, p. 113 n. 1) that only 
universals can be defined and so known in the scientific sense, but 
individuals are recognized by sensation, on which all scientific  
knowledge is ultimately based. He distinguished 'more knowable in its 
nature' (or 'logically prior' Metaph. ioi8b32) from 'more knowable 
to us'. 'By prior and more knowable in relation to us I mean those 
things which are nearest to sense-perception, by prior and more 
knowable in an absolute sense, those which are further from sensation. 
Now the things which are furthest from sensation are above all the 
universals, and those nearest are the individuals.' {An. Post. 7^33- 
72a5.) The theory here refuted by Plato is concerned entirely with the 
sensible world (spoken sounds and musical notes are instances, not 
analogues), and we must suppose it to have taught that just as a 
symphony is composed of single audible units, so ' we and everything 
else' are composed of irreducible physical elements which can be  
perceived but not known.2 If this sounds unsatisfactory, I can only repeat 
my conviction that it is not Plato's invention but an inchoate attempt at 
epistemology by some Sophist or Sophists in the late fifth or early 
fourth century. Whoever invented it, it is an empirical theory like all 
those examined so far. 
Three possible meanings of \ogos Bo6c-2iob). What then must logos 
mean if its addition to true belief is to produce perfect knowledge? 
Socrates sees three possibilities. First, it is the noun from legein (to 
speak), including any expression of thought in words. But this is open 
1 In spite of this, not everyone is convinced that this is all they are. See Bondeson, Apeiron 
1969, 2, p. 7 and A. Rorty, Phron. 1972, 235. But for contrast Hicken, Phron. 1958, 130. 
2 Since much of Aristotle's work is a development or clarification of his predecessors', one 
might compare his distinction between ??????????? (organs like eyes, ears, heart, lungs) and the 
????????? of which they are composed (flesh, blood, bone and so on). The latter he actually calls 
???????? (H.A. 48635, cf. Tht, 205 C7), even though absolutely speaking they are not. 
117 
Parmenides, Theaetetus, Sophist, Politicus 
in some degree to all except the dumb, so that anyone with a true 
opinion would have knowledge as well. This cannot be what the  
statement means. Secondly, to give an account of something may mean to 
enumerate all its parts or elements.1 Hesiod said that a waggon contains 
a hundred pieces. Most laymen would be satisfied if they could name 
five. Knowing, as it were, the syllables but not the letters, they have a 
true belief about what a waggon is but not the complete account (i.e. 
enumeration) which would convert it to technical knowledge. Now if a 
boy spells 'Theaetetus' correctly, one might assume he knows that the 
syllable 'the' is spelt theta epsilon. But the same boy if asked to spell 
Theodorus may begin 'tau epsilon\ thus betraying that in spelling 
' Theaetetus' his correct belief about the first syllable was not  
knowledge, though he listed the elements (letters) correctly. Evidently  
correct belief plus a logos in this sense cannot be called knowledge. They 
must try the third. 
The requirement of knowledge unfulfilled here, though assumed 
rather than stated, is that it must be permanent and unfailing,  
guaranteeing the right answer in every case. A true belief may give one the slip, 
as the Meno said (98a), unless it is secured by the 'bond' or 'tie'  
provided by working out why the answer must be what it is. It must be 
justified true belief. This sense of logos, though common in the phrase 
4 to give a logos \ meaning an explanation or reason, has not been 
considered. 
The last of the three meanings of logos allowed by Socrates is the 
expression of a mark or sign by which the object of enquiry differs from 
everything else, as when we think of the sun as the brightest of the 
heavenly bodies which circle the earth. But here again, the idea of 
knowledge as true judgement plus a logos dissolves under scrutiny. 
Someone has a correct judgement of, say, Theaetetus.2 To become 
knowledge it must, we say, be supplemented by the ability to state a 
mark which distinguishes him from everything and everybody else. To 
1 The idea that complexes can be known and explained but their elements cannot has, after all, 
proved untenable B0566-7). Friedlander (m, 152) identifies the fault in this second suggestion 
with the logical error committed by Tht. himself (i46c-e), as by Meno and others in earlier 
dialogues, of enumerating instances of a universal instead of showing a grasp of the ' one in 
many', 'what is in all instances the same' etc. There is an affinity, but surely also a difference 
between this type of error and that of naming the parts of an individual syllable or material 
object. 2 That is, as usual in these arguments, of who Tht. is. 
118 
Theaetetus 
say that he is a man, and has a nose, mouth, eyes and so on, or even a 
snub nose and prominent eyes, will not do, for it will not distinguish 
him from Socrates and many others. But if we had not already in our 
minds the means of differentiating him from all other men, we could 
not judge correctly who Theaetetus is and recognize him next time we 
saw him. So to add a logos in this sense to true judgement is meaningless, 
for the logos belongs to the true judgement itself, and so cannot be 
knowledge. Nor would it help if we could say that it is knowledge of the 
difference, for to offer ' true judgement plus knowledge of a difference' 
in answer to the question * What is knowledge?' is nonsensical,  
including as it does the definiendum in the definition. 
The definition of knowledge as true judgement plus a logos has 
proved unacceptable on any of the three approved meanings of logos, 
and since Theaetetus has no further definition to suggest, the dialogue 
ends in failure to discover what knowledge is. Its achievement has been 
to rid Theaetetus of several false notions of it, so that if another idea 
comes to him it will be a better one, and if not, the awareness of 
ignorance is always better than fancied knowledge, as Socrates or Plato 
has repeatedly taught in the Apology, Meno and elsewhere. 
As already noted (and often by others) the real—in Plato's eyes— 
relationship between true doxa and knowledge has been omitted. True 
doxa is converted into knowledge by the addition of a logos, in the sense 
of a statement of the reason why it is true, the cause (aitid), as the Meno 
puts it, of its being what it is. It is difficult to believe that the omission 
is accidental.1 The main question throughout has been how we can have 
knowledge of individuals in the physical world—Theaetetus, the sun 
and so on—not of facts, nor of universal concepts like courage or 
justice.2 This is perhaps the strangest feature of the whole dialogue. 
Certainly for Plato the hallmark of knowledge was the ability to say 
4what ? is', i.e. to define it, and we have seen historical and linguistic 
reasons why this should be so (pp. 68 f.). But his ? is always a 
universal or class-concept: justice, courage, or as in the immediately 
1 McDowell's only alternative suggestion (p. 258), that during the actual writing of Thu P. 
lost interest in the definitional task he had set himself at the beginning, is not very plausible. 
2 Moreover the sign or token required is a perceptible one, by which the object can be  
recognized at sight. It is not mentioned that part of the logos distinguishing Tht. from others is that 
he is a mathematician. 
119 
Parmenides, Theaetetus, Sophist, Politicus 
following dialogues 'the Sophist' or 'the statesman', not Prodicus or 
Pericles. Here he spends his time in a vain quest for knowledge of 
individuals, doomed to failure for at least two reasons: 
(i) It could never surmount the difficulty discussed at length by 
Aristotle that definition can only proceed by genus and differentiation 
down to infimae species but not beyond: individuals can never be the 
object of scientific knowledge.1 
(ii) To summarize it yet again, Plato's own doctrine, at once epistemo- 
logical and metaphysical (but are not all epistemologies based on a 
metaphysic, empirical or other?), was that (a) of objects or events in this 
world, where everything is mutable, there cannot be more than true 
belief;2 (b) human reason can classify them and form concepts; (c) from 
the abstraction of concepts a philosopher may proceed to complete 
knowledge of the Forms, which is direct acquaintance, comparable to 
vision. These immutable and intelligible Beings3 are the causes of 
whatever degree of being and knowability physical phenomena exhibit, 
whose status between Being and utter non-Being, corresponding to the 
position of belief between knowledge and complete ignorance (Rep. 
475 ff·? v°l-IV? 4**7 ff-)> they owe to sharing in, or imitating, the Forms. 
Knowledge of individuals, therefore, on any of the hypotheses here 
offered—knowledge as sensation, as true belief, and as true belief plus 
any of the three kinds of account mentioned—was bound to elude the 
searchers. 
Conclusion 
There is little to be said in conclusion that has not been said already. 
The puzzle of the Theaetetus is why, in trying to say what knowledge 
is, Plato should have made little or no use of the doctrine of Forms, 
1 Cf. the passages quoted on p. 113 ?. ? above, and at the head of this ch., p. 61. In arguing 
this Arist. rejects P.'s own definition of the sun as not conveying its essence (?????, Metaph. 
io4oa28-bi). Yet P. was aware of it too. See Allan's introd. to Stenzel's PMD, xxxiif. 
2 That so much is possible, contrary to the extreme flux-theory of sensibles, through their 
relationship to the Forms, has already been explained (pp. 79-82 above). 
3 A Form, though its transcendence makes it individual, retains enough of the character of a 
universal (the ? in things) to be intelligible and definable. Its ambiguous status is the theory's 
Achilles heel, thoroughly probed by Aristotle (e.g. at Metaph. 104038-9). The concepts of 
sharing and imitation he dismisses as 'empty talk and poetic metaphors' {Metaph. 991320). 
When we find these concepts put through a gruelling ex3min3tion in P.'s own Parrn., 3nd in the 
present dialogue a lengthy attempt made to confine the discussion of knowledge to apprehension 
of the phenomenal world, we may be sure that the Forms were being freely discussed and 
criticized in the Academy. 
I20 
Theaetetus 
which even in the critical Parmenides (i35b-c) he admitted to be 
indispensable to philosophical enquiry.1 Has he in fact renounced them? 
I hope I have shown that, mainly in the Digression on the philosopher2 
but also elsewhere, he makes it clear that he has not. Cornford's  
solution, that the Forms are excluded from the main arguments with the 
sole motive of demonstrating the need for them, may be too narrow.3 
As in the Parmenides, Plato is clearly aware of philosophical difficulties, 
and shows an interest in problems, that are absent from the Republic* 
Nevertheless I would not go as far as to say, with Runciman on the 
problem of error {PLE 28), that it is left unsolved because Plato at this 
time 'did not begin to understand the logical and ontological  
misconceptions which underlie the problem as discussed in the Theaetetus9. 
Socrates does not give the impression (especially if one remembers the 
other ostensibly aporetic dialogues) of being a genuinely tentative 
enquirer. His object, as he said, is maieutic, that is, educational. He is 
completely in command all the time,5 drawing out his brilliant pupil 
and then very gently6 indicating the flaws in his answers and getting him 
to consider points which had not occurred to him. The 'usual Socratic 
pose' as an ignoramus, which so enraged Thrasymachus, is well in 
evidence.7 The claims of sense-perception (which ever since the Phaedo 
has provided the first step on the road to knowledge) and of true judge- 
1 The relation of Tht. to the earlier dialogues, and a variety of opinions on its character, have 
been discussed or referred to on pp. 64-6. 
2 Robinson's assertion (Essays 46) that 'The theory of Forms is the theory that there is a 
second world ... and this theory is not implied by the Theaetetus* description of the philosopher' 
is mistaken. What else can the ????? ??????* ????* be A7735)? And what else can ??????? 
?????? ??????? refer to? (i76a-b). On Forms in Tht. see also pp. 102f. above. 
3 Yet so astute a scholar as von Fritz can say (in Essays, ed. Anton and Kustas, 435), with no 
mention of Cornford, that' the Theaetetus tries to show the difficulties into which an empiricist 
theory of knowledge falls when it attempts to do without the Theory of Ideas'. 
4 For a list of these ^ee McDowell 258. But if Cornford's thesis were correct, much more would 
be involved than 'the friere stipulation that the verb "know" is to take only Forms as objects'. 
Robinson similarly oversimplifies when he finds in C.'s interpretation the implication that the 
difference between knowledge and true opinion lies solely in their objects (Essays 56). To use the 
Forms in solving the problem of knowledge involves above all, as C. well knew and the Parm. 
emphasizes, the frightful problem, not of their existence, but of their relation to particulars. May 
Yoh's article 'On the Third Attempted Definition of Knowledge, Th. 20ic-2iob', in Dialogue 
14 A975), is a defence of Cornford's position against Robinson and Ryle. 
5 Campbell anticipated me here (Tht. p. 1): 'He is not himself groping his way. Each footstep 
is firmly planted, as by one who has tried every inch of the path and knows the country well.' 
The echo of the 'road to Larissa', whether intentional or not, is apt. 
6 See 163C5, I99e7, 20531. 
7 Rep. 337a. Cf. Tht. i54C4-d2, i57C7-di. 
121 
Parmenides, Theaetetus, Sophist, Politicus 
ment to qualify as knowledge must be taken much more seriously than 
hitherto, and cannot now be dismissed without a penetrating and 
exhaustive examination; but Plato was never in any doubt that in the 
end they must by themselves fall short of perfect knowledge. 
C) the sophist1 
Introductory remarks. The company of the Theaetetus meet again 
'according to yesterday's agreement', and are joined by a visitor2 from 
Elea who undertakes to characterize three types which are often  
confused but which he believes to be distinct: sophist, statesman,  
philosopher. The question was vital for Plato, and it will not do to say that 
'the argument is hung on the figure of the Sophist quite arbitrarily'.3 
Isocrates would in his view present a living example of the confusion: 
he thought of Socrates and Plato as sophists, and of himself as both a 
philosopher and an authority on politics.4 And the superficial similarity 
between sophistry and Socratic philosophy is emphasized in the  
discovery of a 'nobly-born' variety of sophistry which 'purges the mind' 
of counterfeit knowledge, and which the visitor is doubtful whether to 
call sophistry or not B3od-23ib). It is in fact the Socratic elenchus. 
Again, the Gorgias and Republic have given the impression that  
philosopher and statesman are identical, if one excludes contemporary  
politicians as not 'true' statesmen. In the present group of dialogues Plato 
is looking back more soberly and critically on his earlier enthusiasms, 
and here was a field in which clarification was clearly desirable. Once 
again he employs his unique skill in handling the dialogue form to  
interweave (his own favourite metaphor) more than one theme, and the fact 
that the Sophist uses its avowed subject as a means of advancing also the 
discussion of Being, Non-being and the possibility of error does not 
1 R. S. Bluck, who died in 1963, left an unfinished commentary on the Sophist which was 
published by G. C. Neal in 1975 after this chapter was written. It discusses the opinions of one or 
two scholars whom I have omitted (Kamlah, Moravcsik). 
2 Often called the Stranger, complete with capital letter, which besides its vague suggestion of 
the occult, does scant justice to the word xenos. One translator even imports a Western flavour 
by making Theaetetus begin a speech with ' Well, Stranger . . .' 
3 The quotation is from Edith Hamilton, Collected Dialogues 958, but others have said the 
same. Cf. on the other hand 2i6c-d. 
4 On Isocrates and Plato see vol. iv, 282f., 308-11, and Cornford, PTK 177 with n. 2. 
122 
Sophist 
make it a mere continuation of the Theaetetus. In fact the search for the 
Sophist cannot be separated from this discussion, because the  
arguments to be opposed are all of Sophistic origin. 
In reading the Sophist and Politicus ('Statesman') we must never 
forget that they are only the first two parts of an unfinished trilogy. The 
visitor is to describe three types, and this is repeated at the beginning of 
the Statesman, where Theodorus begs him to take the next two in  
whatever order he prefers, and the visitor replies that he will not give up 
until he has dealt fully with both. Soph. 253 c also looks like a reference 
to the intended Philosopher.1 We cannot therefore expect all our  
questions to be answered in these two dialogues. Indeed it is likely that Plato 
would leave the most important to the Philosopher.2 First he must 
finish his argument with the Eleatics, the opponents most worth his 
steel because they were so nearly right. It was Parmenides who opened 
a window on the truth by introducing into philosophy the notion of a 
changeless, intelligible reality as a prerequisite of knowledge. But their 
absolutism, their 'is-or-is-not' dichotomy with its outright rejection of 
experience, was not a philosophy that could be lived with. A middle 
way must be found between this and the doctrine of total flux, and for 
that no better discussion-leader could be found than one brought up in 
the school who is yet an independent thinker. With the difficulties 
about 'is-or-is-not' removed, the way would be clear for a positive  
restatement of Plato's own conception of knowledge, and it is a most 
attractive conjecture (it cannot, alas, be more) that for this the leading 
role was to be restored to Socrates himself.3 Meanwhile hints are given, 
1 These passages leave no room for reasonable doubt that P. planned the Philosopher. Most 
scholars agree, and conjecture that he was prevented from writing it either by the current of his 
own thoughts (Cornford, PTK 323) or by circumstances such as his last visit to Sicily and 
consequent disillusionment (Wilamowitz, PL 1, 558; Leisegang, RE 2354k). Friedlander 
however (PI. 111, 281; 525 n. 5) thought the project impossible and P.'s references to it ironic. 
The Sophist is a counterfeit philosopher, but the true statesman is the philosopher himself. (I do 
not find this plausible.) Since antiquity unsuccessful attempts have been made to identify it with 
an existing dialogue. See Taylor, PMW 375 n. 1. 
2 Cf. p. 154 below. Wilamowitz maintained (PL 1, 559k) that the lack of the Philosopher has 
robbed us of what was to P. the main point, and caused scholars to assume that he had given up 
much which, because of its importance, he had reserved for the concluding dialogue. Thus of the 
two problems left unsolved in Tht.—the definition of knowledge and the possibility of error— 
the first is not taken up in either Soph, or Pol. because knowledge is the prerogative of the 
philosopher. 
3 At Pol. 258a S. proposes that his young namesake should be respondent, now to the visitor 
and later to himself, which Cornford (PTK i68f.) thought difficult to explain except on this 
123 
Parmenides, Theaetetus, Sophist, Politicus 
as in the Digression of the Theaetetus, of the sort of man the philosopher 
is. The visitor, as a philosopher himself, is, * though not a god, godlike; 
genuine philosophers look down from a height on the life of those 
below' Bi6b-c); and whereas the Sophist takes refuge in darkness, the 
difficulty in discovering the philosopher lies in the very brilliance of the 
realm of reality in which his mind dwells, 'dark with excessive bright', 
for the eyes of ordinary souls cannot endure to look steadfastly at the 
divine B54a-b). 
At 217c the visitor is given the choice between continuous  
exposition and question-and-answer, and chooses the latter, provided the 
respondent is 'docile and gives no trouble'. There is, then, to be no 
genuine argument, but the retention of spoken dialogue does permit 
the enlivenment of much dry logical argument by the humour,  
metaphors and other characteristic touches which we have come to expect of 
Plato.1 
Definitions: the angler and the Sophist Bi8e-3ie). The visitor and 
Theaetetus (his chosen respondent) both use the word 'Sophist' but do 
not yet know whether they have the same idea of what they mean by it 
B18 c). The Sophist is in fact an awkward creature to track down, and 
the visitor suggests they try out his proposed method first on  
something simple and trivial, say an angler. With no preliminary explanation 
of what this method is, no laying down of principles or rules, he plunges 
at once into the demonstration by example, from which we see that the 
first step is to fix on a very wide class in which the subject can safely be 
included. No one will question that the angler practises an art (techne). 
Arts are then divided into two, in this case productive and acquisitive, 
with angling assigned naturally to the acquisitive branch, and  
acquisitive into peaceful (e.g. by persuasion, barter, purchase) and forcible. So 
the dichotomies continue, expressible in a stemma in which the right 
hand member is always chosen and the left discarded, until the subject 
supposition. Of course the visitor has promised to deal with the philosopher too, but as in Tim, 
A7 a) P. could easily invent a reason for his absence on a later occasion. 
1 It may amuse readers to check their own impressions with those of Thompson in the long 
paragraph quoted by Campbell (ed., p. xliii), which ends: 'If vivacity in the conversations, easy 
and natural transitions from one subject to another, pungency of satire, delicate persiflage, and 
idiomatic raciness of phrase are elements of dramatic power, I know no dialogue more dramatic 
than the Sophistes.' 
I24 
Sophist 
is defined by the original genus and a 
? ' 1 consecutive series of differentiae.1 By 
| I this method the angler appears as a 
| I practitioner of acquisitive (not  
productive) art, forcible (not peaceful), by 
hunting (not fighting), of animals (not the inanimate) and water (not 
land) animals, fish (not fowl), by striking (not netting), with hook 
(not spear). 
Not only is the method clumsy, but some of the divisions are  
questionable; all birds are apparently classed as the winged (as opposed to 
underwater) division of swimming animals. And though learning is  
undoubtedly the acquisition of knowledge, one wonders a little about a 
method that classifies it with trade, fighting and hunting as an  
unproductive art engaged in laying hands on what has been produced or 
preventing others from laying hands on it B19c). But before  
attempting a judgement let us see how it is applied to their real subject, the 
Sophist. Here a coincidence strikes the visitor: the Sophist is a kinsman 
of the angler, for he too is a hunter of animals. At this point however he 
diverges and takes up the left-hand alternative which he discarded in 
defining the angler; for the Sophist is a hunter of land animals (not 
water), tame (not wild). (Here he has to pause to win agreement from 
Theaetetus that man is a tame animal and is hunted.) After further 
dichotomies the Sophist's art is defined as acquisitive, hunting tame 
land animals, viz. men, privately for money, capturing youths of wealth 
and reputation under the pretext of education. 
The definition is hardly a model of objectivity, and together with the 
pretended discovery that Sophistry is a kindred art to angling,  
ostensibly chosen at random simply to illustrate the method, shows from the 
start that, whatever Plato's opinion of the value of diairesis in general, 
what he is giving us here is satire not philosophy. Success in the method 
would demand a thoughtful and unbiased choice of genus and  
successive differentiae, not one designed to show the object in the worst 
possible light. In fact, with no other excuse than that the Sophist's art is 
1 A full table is given by Campbell (p. 24), Taylor (PMW 378), Ritter {Essence 238) and 
others. Bluck tabulated this and the other diaireseis in the dialogue on pp. 55-7 of his  
commentary. 
125 
Parmenides, Theaetetus, Sophist, Politicus 
formidably complex, six further definitions are given, each starting from 
a different genus. One metaphor retained throughout is that they are 
'hunting' the Sophist, who, Proteus-like, in trying to escape their nets 
will assume all sorts of forms. By adopting different starting-points he 
is revealed in seven different guises. The first we have seen. Here are the 
next five.1 
B)-D) B23c-24e). These are in fact three further varieties of the 
Sophist in his character as money-maker. Taking up the other main 
branch of acquisitive art, by peaceful exchange, and following it 
through three series of subdivisions, we find he is 
B) An inter-city trader in food for the mind concerned with the 
learning of virtue. 
C) A retailer of the same wares in his own city. 
D) A manufacturer of them for sale.2 
E) B25a-26a). Returning to the other branch of acquisitive art, 
acquisition by force, and that half of it which consists in open contest 
not stealth, we continue subdividing until we find the Sophist to be a 
pugnacious debater or eristic. 
F) B26b~3ib). For this definition the original dichotomy of arts 
into productive and acquisitive is abandoned, and a new widest genus 
is selected, the arts of separation, in particular those which separate 
worse from better and so purify. Purification may be of body or of 
psyched Of the latter, the most important is that which purges the error 
of believing one knows what one does not, and for this the most 
1 In general I have followed P.'s own summary at 231 d-e rather than the details given earlier. 
Cornford (PTK 187) saw this as really a classification of Sophists, each definition referring to a 
different person or group, but this can hardly be maintained. He put Protagoras in the rhetorical 
group as distinct from the agonistic or eristic type represented by Euthydemus, and Hippias 
among the 'teachers of advanced subjects'. But Protagoras was a veteran in ?????? ????? (Prot. 
335 a) and Hippias was undoubtedly a thetorician. See vol. in, 44k (following H. Gomperz and 
E. L. Harrison in Phoenix 1964, i<)of.), and for Hippias ib. 280 ff. But in any case rhetoric is 
excluded from Sophistic in this dialogue. See pp. 157 k below. F. Oscanyan (Philos. Forum 
1972-3, 241-59) has made out an interesting case for assigning each definition to an individual, 
namely Gorgias, Protagoras, Hippias, Prodicus, Euthydemus, and finally (with qualifications) 
Thrasymachus as the 'nobly-born Sophist'. 
2 The notion of the Sophist as a seller of mind-food is repeated from the Protagoras C13 c). 
3 Here P. divides the impurities of the psyche into wickedness and ignorance, corresponding to 
disease and deformity in the body, and to be cured by punishment and instruction respectively. 
Cornford (PTK 179) regarded this as going beyond S., for whom wickedness was identified with 
ignorance. However, in the Laws (86od, pp. 376-8 below) P. emphatically re-asserts the Socratic 
dictum that all wrongdoing is involuntary, but then argues with some subtlety that this does not 
126 
Sophist 
efficacious method is not chiding and exhortation (for all such ignorance 
is involuntary) but the elenchus, which by revealing inner  
contradictions makes a man aware of his own ignorance and prepared to receive 
positive teaching. To call one versed in this art a Sophist is questionable, 
but Met it pass' B3138-9). (More of this below.) 
Comment on definitions 1-6. The production of multiple definitions by 
assigning the subject to different genera in turn may perhaps be justified 
in terms of general method (it is defended by Taylor, PMW379), but 
Plato himself upholds it on the special ground that their quarry in this 
particular case is a slippery and many-sided creature who gives the 
appearance (????????) of belonging to more than one class.1 The 
genera chosen without question—hunter, money-maker, dealer in 
unrealities—show the satirical, pseudo-scientific character of the  
exercise. Plato evidently had no individual Sophist in mind, but a  
combination of all that he found objectionable in the Sophistic profession. Only 
the sixth stands apart, and on this and no. 5 a historical note may be in 
place. 
E) In his summary of this definition at 231 d-e Plato stops at eristic, 
but earlier B25 d) he had added a further dichotomy: of eristics, one 
sort (the Sophist) debates for money, the other for its own (not its 
hearers') pleasure, neglecting its fortune to do so. These should be 
called chatterers or babblers (?????????). Who are these? Cornford 
{PTK 176) chose the Megarians, against Campbell who thought 
Socrates himself a possibility. I should call it a certainty. Socrates  
impoverished himself in his zeal for the elenchus {Apol. 23 b, 3ib-c), his 
partners in argument certainly did not always enjoy it as much as he 
did, and his detractors called him * chatterer' (?????????), a term which 
Plato defiantly adopted as the badge of true philosophy.2 Cornford 
objected that he would not call Socrates an eristic, but that too he was 
called by others, and in the same ironic spirit Plato could enjoy a covert 
for practical purposes invalidate the distinction between voluntary and involuntary  
misdemeanours as commonly accepted nor abolish the need for punishment, which may be the best 
cure for what is in fact a disease of the psyche (862 c, as here at 222 b 8). 
1 223c; ???????? again at 226a6, ????????????? 231 a8. Cf. ????? ???????? 231 b-c. 
a For reff. see Cornford, PTK 176 n. 3 (where however 'Statesman 270a' should be '299b'), 
and vol. iv, 431. 
127 
ParmenideSy Theaetetus, Sophist, Politicus 
reference to him as an eristic of a different stamp from the Sophist. The 
point is not so much that he is called an eristic as that he is sharply 
separated from the Sophist, with whom he was so often confused. In 
the early dialogues Plato frankly demonstrated his use of eristic 
techniques.1 
F) Here the visitor gives a precise and detailed description, not of 
Sophistry but of the elenchus as practised by none but Socrates himself. 
To confute the few who have thought otherwise,2 one need only quote 
the passage at some length. 
B30 b 4 ff.) They [sc. the purgers of the mind otherwise than by admonition] 
cross-examine a man on a subject on which he thinks he has something to say 
though really saying nothing, and since such people are all at sea, have no 
difficulty in exposing their beliefs by putting them side by side in discussion 
and showing them to be mutually contradictory. When the speakers see this, 
they become vexed with themselves and more tolerant towards others. Thus 
they are delivered from pretentious and obstinate opinions in the way of all 
ways most enjoyable to the listeners [cf. ApoL 23 c, 33 c] and of lasting benefit 
to themselves. Their purgers treat the mind as doctors the body. As doctors 
hold that the body can get no benefit from food until some inner obstruction 
is removed, so these consider that a man's mind will not profit by an intake 
of knowledge until someone refutes and shames him, and by ridding him of 
the mental obstacles to learning purifies him and leaves him persuaded that he 
knows what he knows and no more . . . For all these reasons the elenchus 
must be called the greatest and most effective kind of purgation, and he who 
has not undergone it, be he the Great King himself, is in the highest degree 
impure, ignorant and foul in the very respects in which to be genuinely happy 
a man ought to be purest and fairest3 . . . What shall we call those who 
practise this art? Personally I shrink from saying * Sophists'4... Nevertheless 
1 E.g. in Charm, and the Hippias dialogues. See esp. vol. iv, 159, 185 and 186, 195 f. 
a See the controversy between Kerferd and Trevaskis in CQ 1954 and Phron. vol. 1, 1955, 
in which T. had the better of the argument. Burnet rather absurdly thought of the Megarians 
G*. to P. 276). The disguise of the plural can be disregarded. 
3 Cf. Gorg. 470 c This and the one mentioned at p. 126 n. 2 are of the sort of echo that scholars 
use in other cases as evidence that a dialogue is spurious, 'put together by a botcher'. 
4 Here follow the words ?? ?????? ?????* ??????????? ????* (' for fear of doing them too 
great honour'). The antecedent of ?????* is debated. Jackson, Cornford (PTK 180 n. 2) and 
Taylor (PMW 381 n. 1) thought it could not be the Sophists, and must be said ironically of S.: 
he only sought knowledge (was ?????????) and cannot compete with the man who has it (the 
wise man or ????????). Kerferd and Trevaskis (CQ 1954, 85; Phron. 1955, 37) revived the 
contrary view, which is both possible and likely. In these descriptions of the Sophist P. uses the 
bludgeon rather than the rapier, and the elenchus, he goes on to say, can only be called Sophistry 
if we admit that it is a superior brand. 
128 
Sophist 
in our present discussion1 let the elenchus of fancied knowledge be simply 
called * Sophistry of noble extraction'.2 
The elenchus as here described was not the procedure of Euclides 
(vol. hi, 506), still less of the arrogant young followers of Socrates who 
brought his name into disrepute (Apol. 23 c). It is an idealized picture 
of his own method and its effects, as described in the Apology and seen 
in action in many a Socratic dialogue—idealized because, unfortunately, 
in real life the adult and opinionated grew angry with him instead of 
themselves (ApoL 21 d-e, 22e-23 a). Only initially modest and receptive 
young men like Charmides or Theaetetus could benefit from Socratic 
psychiatry. Indeed the parallel between this passage and the closing 
words of the Theaetetus is strikingly close. All his life Plato had in mind 
the tragedy that Socrates was commonly ranked with the Sophists. The 
inclusion of his elenchus among descriptions of Sophistry makes it 
stand out in vivid contrast to the rest, and Plato says in effect: * Call it 
Sophistry if you like, but then " You and I have only the name in 
common, not the reality" [cf. 218c], unless we agree to include under 
Sophistry something entirely different in its aims and results from any 
other form of Sophistry, and of an altogether higher status; and 
personally I would rather not.' 
Diairesis^ It is convenient to use this term, which is simply the Greek 
for * division', for that division into kinds which was a part of Platonic 
method. Later in the Sophist he says B53 d, exactly as at Phdr. 265 d ff.) : 
To divide according to kinds, not mistaking one fornvt for another, belongs 
to the science of dialectic. Whoever is capable of doing it distinguishes one 
1 LSJ offer no encouragement to follow Campbell and Cornford in translating ?????????? 
' appeared by a side wind*. It simply means to appear or turn up. Cornford himself translates it 
'coming in sight' at Tht. 199c8, and cf. Ar. Poet. 144932 ???????????* ??* ????????*. 
a Or 'of a noble kind'. English cannot reproduce the double reference of ????*: (?) family or 
descent, (b) kind, genus. Mortley's remark (Eranos 1969, 30) that 'it is difficult to see why a genus 
should be called noble' ignores this. His argument also suffers from identifying ???????* with 
??????*. 
3 In a lucid and helpful account, Ackrill has defended against Ryle the standing of diairesis as 
'a significant part or instrument of philosophy' both in P.'s mind and in fact. See his 'In Defence 
of Plato's Division' in Ryle> 373-92. Diairesis is taken up again in connexion with the Politicus 
on pp. i66ff. below. 
4 In using the small initial here I do not wish to prejudge the question whether the ???? or 
???? referred to have the status of Platonic Forms. It should be noticed that ????*, ????* and 
???? are used interchangeably, as are ????? and ????* at Pol. 262d, not to distinguish genus from 
species. 
IO 129 GHG 
ParmenideSy Theaetetus, Sophist, Politicus 
form everywhere extended throughout many,1 each of which lies apart, and 
many forms different from each other embraced from without by one; and 
again one form unified through many wholes, and many in every way distinct 
and apart.2 This means knowing how to distinguish, kind by kind, how the 
several kinds can and cannot combine. 
This dialectical skill, he adds, is the philosopher's. 
This complex process, going beyond the simple dichotomous  
classifications of the early part of the dialogue, is the climax of a fairly long 
development, originating in the Socratic search for definitions (pp. 27 f. 
above). That was carried out by dividing things according to their 
kinds' (Xen. Mem. 4.5.12), e.g. in defining courage by separating it 
from misplaced stubbornness or from rashness {Laches). An early form 
of the Platonic procedure also appears in the Socratic Euthyphro 
(i2dff.), where to discover the nature of piety the whole field of right 
conduct is divided into conduct towards men and conduct towards 
gods. A more elaborate division is carried out in the Gorgias D64 b if.), 
which Dodds sees as exemplifying the method of Sophist and Politicus, 
but (following Cornford) calls *a Platonic, not a Socratic invention'.3 
It recurs in the Republic D54a), where eristics are distinguished from 
dialecticians by their inability 'to divide according to kinds'. The first 
formal description of the dialectical method occurs at Phaedrus 265 d-e, 
though earlier in the dialogue B49 b) Plato has mentioned the universal 
human power of forming a general concept from a mass of individual 
sensations. First the dialectician, taking a synoptic view, brings many 
scattered kinds under a single generic form, including the subject to be 
1 I am in doubt whether to follow Cornford in taking ???* ??????? to refer to Forms, or 
Runciman (PLE 62) who, observing acutely that if this were so, grammar would require 
???* ??????*, concludes that particulars must be meant. Yet the whole context here is concerned 
with the relationship of Forms, which of them can and which cannot combine, and the rest of the 
sentence certainly suggests that ?????? at e ? are Forms. Plato could be thinking automatically of 
the terms ???? and ???? which he has just used (di), and commonly uses in this dialogue in 
preference to the quite exceptional ????. (Now that Bluck's Sophist has appeared, I see that he 
agreed with Runciman. Cf. his p. 127. On pp. i3of. he also offers a different interpretation of 
ei-2, taking ?????? to be particulars.) 
* Bluck {Soph, 127-31) criticizes several interpretations of the last part of this sentence, and 
offers his own. 
3 For criticism of this view see p. 28 n. 1 above. I am puzzled by Luce's remark (CQ 1969, 
230) that Socratic definition puts forward a name and asks for its logos, whereas much of the 
procedure in the diaireseis of Soph, consists in putting forward a logos and asking for its name. 
Does it not put forward the names 'angler' and 'Sophist' and ask for their logoi? 
I3O 
Sophist 
defined (in this case love), thus marking it off* from the members of other 
genera.1 Secondly, the generic form is carefully divided 'at the natural 
joints' until by applying successive differentiae the infima species 
('indivisible', Phdr. 277by, Soph. 229d5) is reached.2 Formally  
therefore the dialectical process is a double one, a collection or bringing 
together' (???????? Phdr. 266b) followed by diairesis, and to this 
method Plato was faithful all his life. Not only does he honour, as well 
as practise, it in the Politicus (see especially 285 a-b, 286 d), but in the 
Laws (965 c) he was still writing that there is no clearer or more  
accurate way of investigating anything than by pressing on from many 
different instances to a knowledge of the single form, and then ordering 
them all in relation to it. The first stage is not mentioned in the Sophist 
(which is singularly lacking in explanation of the method it employs), 
and little use is made of it. In the sixth definition Reparative art' is 
reached through the mention of various homely processes—filtering, 
sifting, winnowing, * combing' (in weaving)—but in the others the 
generic form is taken to be self-evident. In the elaborate statement of 
Soph. 253 d, the term diairesis seems to include the preliminary process 
of collection. 
In the early part of the Sophist, Plato writes as if dichotomy were an 
integral part of the method, but as a classificatory tool its usefulness is 
obviously limited, and elsewhere he speaks of it only as preferable but 
not always possible;3 Aristotle, in his work On the Parts of Animals 
1 Some have thought particulars are meant (or at least included: see Hackforth, PEP 142f.). 
But the method being described is purely the dialectician's B53d2-3), whereas generalizing from 
particulars is a universal human accomplishment. The generic form is that One form, the same 
in all', which S. wants Meno to identify in Meno (vol. in, 433 n. 1). There too the 'many' 
falling under the one form are universals (kinds of virtue), not ' individual things'. Contrast 
Cornford, PTK 185, 186; but the 'Socratic muster' was never of individuals. (He accepts Meno 
as Socratic, 184 n. 2.) 
a Arrangement of these volumes by dialogues has the drawback (for which I hope the  
compensations are adequate) that it necessitates either repetition or the inconvenience of cross- 
references. P. of course chose repetition. I have attempted compromise. For Socratic diairesis see 
vol. in, 440, and vol. iv, 431 n. 1, for Phdr. vol. iv, 427-31, and for an outline of the method of 
diairesis vol. iv, 47 f. 
3 Soph. 229a-b, instruction has 'more than one' kind, but two are especially important; Pol. 
286d, if dichotomy is impossible one must (as in Phdr.) divide limb from limb 'like a sacrificial 
victim'; Phil. i6d, divide the one form into 2, or if necessary 3 or more. In Phdr. divisions into 
2,3 and 4 are used. (See tables in vol. iv, 429 f.) That diairesis was eagerly carried on and discussed 
in the Academy is obvious from the pages of Aristotle. Speusippus is thought to have been one 
who defended dichotomy as by itself sufficient for classification. Details in Cherniss, ACPA 27-63; 
more briefly Skemp, Politicus 70-3. But see also p. 464 n. 5 below. 
131 
10-2 
ParmenideSy Theaetetus, Sophist, Politicus 
(bk i, ch. 2), argues in detail against it as a method of classification in 
biology. Biological distinctions figure in Plato's definition of the angler,1 
and since we know that zoology and botany were zealously pursued in 
the Academy, it is probably in this connexion that the method of 
diairesis was developed. The parody by the comic poet Epicrates is well 
known, in which he depicts Plato, Speusippus and Menedemus teaching 
pupils to * divide' animals, trees and plants and setting them to assign 
the pumpkin to its proper class. Of Speusippus we have a number of 
quotations from a meticulous work on classification and nomenclature, 
in which for instance he groups no less than six species of bivalves 
together, then oysters and limpets in a different sub-group, and  
distinguishes four species (????) of polyp.2 
jaeger's claim that biological classification was pursued by Plato only 
'in order to learn the logical relations of conceptions' {Arisu 19) seems 
to confuse Plato's appreciation of the truth that science can never  
penetrate below the level oiinfimae species (which for him was based on the 
doctrine of Forms) with a complete lack of interest in the sensible 
world. The mode of existence of individuals, and the nature of our 
cognizance of them, were always in the forefront of his thoughts, as the 
Theaetetus has shown. (See further pp. 412-17 below.) This brings us 
to another much-discussed question: how seriously are the dichotomies 
of the Sophist intended? Leisegang {RE 2493) found the paradigm 
definition of the angler so capricious and absurd that it must be simply 
a mockery of the whole procedure. Apelt {Soph. 30 f.) wanted to  
distinguish ridiculing the Sophist from ridiculing the method, and even 
held that the reason for offering half a dozen definitions was to give the 
reader plenty of illustrations of a procedure of which Plato thought so 
highly: it was a gross error to suppose that he would laugh at the method 
itself. That the divisions are biased and polemical is obvious. That they 
display wit, jest and lightheartedness is Apelt's own admission. Beyond 
that each must judge for himself, but it is at least possible that Plato is, 
as one might put it, being his own Epicrates, and having a little fun at 
1 Arist.'s objection to classing some birds as water-animals F42b 10-13) looks like a reference 
to Soph. 220 a-b or some Academic scheme on which it is based. 
2 Speus. frr. 8 and 16 Lang, pp. 463 f. below. On pp. 8-15 Lang notes the close affinity between 
his classifications and Aristotle's. The Epicrates fr. comes from Ath. 2.59 (fr. 11, p. 287, Kock, 2, 
354 Edmonds). Further reff. in vol. iv, 22 n. 2. 
132 
Sophist 
the expense of over-enthusiastic colleagues who were advertising 
diairesis, especially in its dichotomous form, as the universal key to the 
problem of knowledge. There may also be an element of self-criticism, 
as in the later argument against the * friends of Forms' (pp. 141-3 
below), for having spoiled a fundamentally sound thesis by over- 
extending its field of application. 
The method of diairesis has often, and rightly, been praised as the 
foundation of scientific classification, and no doubt it was Aristotle's 
experience in the Academy, as well as his natural bent, that set him on 
the road to becoming a biologist, superior in Darwin's eyes to Linnaeus 
or Cuvier. Its usefulness extends also to mathematics, a subject nearer 
to Plato's heart. But as a general philosophical method it perhaps bears 
too clearly the marks of its inheritance in its conception of all philosophy 
as comprised in an answer to the Socratic question of what a thing is, 
culminating in the majestic doctrine of the objectively existing Forms 
as the explanation of all being and knowledge alike. As we have seen 
in the Theaetetus* knowledge for Plato would always present itself as 
knowledge of some * thing' rather than * knowledge that' or * knowledge 
how'. Moreover to grasp the one Form above the many is not simply 
the last stage in a process of thought but an achievement of direct 
acquaintance with the divine world in an act analogous to vision. This 
for many is the core of Platonism, and for this reason (to voice an 
unpopular view) his greatness may be thought to show itself most clearly 
in the dialogues written in full assurance that this was the truth, before 
the pristine vision was clouded by doubts—which in any case never led 
him to abandon the assimilation of all knowledge to knowledge by 
direct acquaintance of * what is'. 
Seventh and Final Definition: the Sophist as Illusion-maker {nominally 
from 232 b to the end). The first six diaireseis have really only revealed 
six aspects, or manifestations, of our elusive subject. To get him wholly 
in the net, says the visitor (following the procedure described in Phdr. 
265 d but practised by Plato since the earliest Socratic dialogues), we 
have to find the common element in all of them. As always, the Socratic 
assumption is taken for granted, that a common name implies a common 
1 Cf. esp. pp. 67-9 above. 
133 
Parmenides, Theaetetus, Sophist, Politicus 
nature. To begin with, all set up as controversialists and teachers in 
controversy on any subject whatever, even writing books which  
profess to outdo an expert in his own field, from theology to physical 
science, politics or even wrestling.1 Since no man can know everything, 
what they offer their pupils must be the appearance, not the reality, of 
knowledge. As an artist might deceive young children, if they were not 
allowed to approach too close, into thinking that a painted scene was 
real, so the Sophist exhibits simulacra in words (?????? ????????, 
234 c), which those far enough removed from the truth mistake for 
realities, and ascribe all wisdom to their authors. 
So the Sophist is placed in his genus: he is an illusionist or imitator2 
of reality B346-353), and applying diairesis the visitor discerns two 
kinds of imitation. A modeller may either reproduce the original 
exactly, in its proper dimensions and colouring, or intentionally distort 
its proportions (as in designing an over-lifesize statue for a high  
building) to make them appear correct from a particular distance and  
viewpoint,3 though if seen close at hand and at eye-level the deception would 
be obvious. The reader may feel pretty sure which division the Sophist 
will end up in, but at this point the division is broken off and not  
resumed until near the end of the dialogue, for in the character of imitator 
their cunning quarry has found a dark and obscure hiding-place. Just as 
in the Theaetetus the definition of knowledge as true belief raised the 
whole question of the possibility of falsehood, so here before one can 
brand the Sophist as imitator one must settle the same baffling problem: 
c This appearing and seeming without being, and the saying of things 
but not true things, are now as in past time thoroughly perplexing. 
How one may say that there really is such a thing as false speech or 
belief, without being caught contradicting oneself, it is very hard to 
see' B36c). 
1 For wrestling Tht. mentions Protagoras. Cf. also the claim ascribed to Gorgias that the 
rhetorician's skill in persuasion could get him the job of public physician in competition with a 
doctor {Gorg. 456b-c). The criticism of the Sophist as offering ????????? ???????? also goes 
back to Gorg. Cf. 459ci, ?6 (?????? ??????? ??? ????? ?? ??? ??????), e5· 
* As has often been said, P. obviously intends to recall the derogatory description of mimesis 
in Rep. 10 (vol. iv, 545 ff.). Cf. esp. 233d~34a with Rep. 596b-e. Many other parallels are pointed 
out by Dies, Bude* ed. 271. 3 Cf. Rep. 602 c-d. 
J34 
Sophist 
The status of'what is not' and the criterion of being B37 a—48 e) 
(a) The Sophist's reply B37b~4ib). To believe that falsehoods can 
arise is to believe that 'what is not is', which 'the great Parmenides' 
expressly ruled out as impossible. First, then (to 241b), the visitor 
simply states the Eleatic case, and his language, besides one direct 
quotation from the' Way of Truth', is steeped in Parmenidean  
phraseology. 'What is not' is confined to 'what in no way is', i.e. the  
absolutely non-existent, for so Parmenides understood it. We utter this 
phrase, but to what can it refer? ' What is not' can be neither one nor 
many nor have any attributes at all, for if it had, it would in some way 
be. Even to deny 'its' existence is to call it singular. It simply cannot 
be thought or spoken of at all, as the great man said. 
If, then, we call the Sophist a maker of verbal 'images'1 he will 
immediately ask what we mean by an image, forcing us to say that it 
really is an image, but is not the 'real thing' whose image it is, and so 
contradict ourselves by saying that 'what is not in some way is' 
B40C4-5). Moreover, if we accuse him of deception, we can only mean 
that he induces false beliefs in others, and he will not fail to point out 
once more that to believe a falsehood is to believe that what is not is, 
which we have just agreed to be impossible. 
The use of the Parmenidean dilemma for eristic purposes was a 
genuine mark of the Sophists. Plato had not forgotten the declared 
purpose of the dialogue, to get the measure of people like Euthydemus 
and Dionysodorus. But what he attributes to them here contains a few 
curious features. Theatetus thinks the Sophist's imagined question— 
What do you mean by an image?—an easy one. 'Obviously we shall say 
we mean images in water and mirrors, also pictures, models and so on.' 
But the Sophist will refuse to look at these visible objects and demand 
an answer based on logoi. He will want to know (to translate 24034-6 
as literally as possible) ' what permeates all these many things in men- 
1 ??????????? 239 c! 3. An ??????? was anything which gave the appearance of something 
without being the thing itself—ghost, reflection, painting or statue. P. coupled it with 'false' at 
Tht. 150c, and it is usually synonymous with ????? (cf. Rep. 509? with Soph. 239d6-8), though 
at 239 d 3 it appears, in the light of the Soph.'s division between ??????? and ??????????, to 
stand for the latter, deceitful type of imitation. This division is of course ad hoc. If at Rep. 
509e-ioa P. calls reflections ???????, at 516b ???????????? andat&pA. 239d ??????, that only 
illustrates his dislike of a technical precision of language (Tht. 148c). 
135 
Parmenides, Theaetetus, Sophist, Politicus 
tioning which you thought fit to call them by one name, "image", 
covering them all as a unity'. Now (i) this is the standard Socratic 
procedure for eliciting a definition,1 which when it comes (in this case 
'another thing of the same sort copied from the real thing') is shown to 
be untenable as in several Socratic dialogues. Theaetetus, it appears, has 
not profited by the lesson he was given on the previous day. (ii) It is a 
method of which, in the Hippias dialogues, a Sophist is proved to have 
no understanding, (iii) Stranger still, to refuse to look at visible things, 
and seek the truth about reality in logoi, is precisely what Socrates 
describes himself as doing in the Phaedo (99d-e) when, baffled by 
earlier and contemporary efforts to discover reality and its causes within 
the sensible world, he postulated the doctrine of Forms. 
The argument is genuinely Socratic, not a parody.2 Plato (I surmise) 
found it amusing to make the so familiar point ex persona Sophistae, 
with the subtle justification that the visitor is still assuming the purely 
Eleatic premise, on which so many Sophistic paradoxes were based, that 
there is no third choice between 'is' and 'is not'. The full subtlety and 
skill of the composition are revealed when the Sophist has the ground 
cut from under his feet by the acceptance of what he called impossible, 
that' what is not in some way is, and what is in a way is not' B41 d). To 
defeat the Sophist Parmenides himself must be called in question. 
Another point should be noted. The diairesis was broken off because 
the visitor could not decide which of the two species of image to ascribe 
to the Sophist's art, the (as near as possible) replica or the distorted-for- 
effect, the eikon or the phantasma. (Cornford's terms 'likeness' and 
'semblance' are convenient.) In what follows, however, it is simply 
assumed that he is a maker of semblances,3 and when the diairesis is 
resumed he is at once placed in that division without comment or  
question B66e-67a). It is hardly fair argument, just a reminder of Plato's 
ineradicable conviction of the harmfulness of the Sophistic art. 
1 The phrases ?? ??? ?????? and ??? ????? both occur in Meno G4a and 75 a) when S. is 
trying to get Meno to see the same point. 
* Campbell ad loc. calls this passage a caricature of Socratic method, but though the Sophist's 
feigned blindness is amusing, the method is not caricatured. The imperturbable gentleman' of 
Tht. 165 c is irrelevant, and C. does not mention the striking parallel with the Phaedo, 
3 239C9, 24odi. P.'s incorrigible distaste for a fixed terminology appears again when, even 
after the diairesis of ??????????? at 235C8-36C7, he uses ????? as equivalent to ??????? in 
general (i4obn-i3). 
I36 
Sophist 
(b) From the unreal to the real B4213-45 e). After declaring that he must 
lay unfilial hands on his own * father' Parmenides, the Eleatic does not 
immediately do so, but makes a fresh start. Since the notion of the  
unreal ('what is not') has led to perplexity, let them turn to the real. Are 
they so sure they know what they mean by that?1 What have previous 
thinkers made of it? The natural philosophers talked of one or more 
'real things', e.g. two physical opposites like hot and cold or wet and 
dry, or said that what is was both one and many, either successively or 
even both together. Whether right or wrong, they treated us like 
children, speaking in mythical terms of these real things as moved by 
hatred or affection, fighting, marrying and begetting. The Eleatics, 
going back at least to Xenophanes, wove their myth on the theme that 
all things (so called) were One.2 None of them made their meaning 
clear. Those who declare that there are two realities, say hot and cold, 
must be asked what is this being (or reality) that they attribute to them3 
severally and together. Is it something else besides them, making three 
in all? They cannot identify it with either separately, for then there 
would be one reality, not two; but if they identify it with both, that too 
is to reduce both to one. The monists are in no better case. Is being the 
same as one} How can there be two names on the monist hypothesis? 
How can a name exist anyway? If it is different from the thing named, 
they are two. If not, it is the name of nothing. Again, it must be a whole 
of parts, for Parmenides compared it to a sphere with centre and  
circumference.4 As sum of its parts such a whole can have unity of a sort, 
but cannot be 'the One' itself. 
This rather strange language has been explained by Cornford {PTK 
1 As others have pointed out, the question 'What is Real?' is not answered in the Sophist. For 
P. as for Aristotle it was the fundamental question of philosophy, and it is reasonable to suppose 
that it was to be dealt with in the Philosopher. 
2 Only Xenophanes is named, and P. is not attempting a history of Presocratic thought. But 
one can recognize the Eros of Hesiod and the Orphics, and the biological analogies of the early 
Ionians, as well as the Love and Strife of Empedocles and the antinomies of Heraclitus. The 
proponent of a triad of beings B42 c 9) could be Pherecydes. See fr. 1 DK. 
3 Only the ultimate ????? were recognized by the physikoi as being (????). Other phenomena, 
as derivative and transitory combinations of these, had no existence of their own. 
4 The visitor quotes fr. 8 11. 43-5 DK. Plato next considers the consequences if'what is' is 
not a whole, but since it is agreed that the Eleatics do say it is a whole, we may perhaps spare 
ourselves his intricate reasoning on this point. It is analysed by Cornford, PTK 223. Taylor 
(PMW283 n. 2) and Schofield (CQ 1974,42) point out how this short section resumes arguments 
from the Parm. 
137 
Parmenides, Theaetetus, Sophist, Politicus 
222 f.), though its point need not depend on the existence of Platonic 
Forms, of which Cornford makes much (and they were of course in 
Plato's mind). Earlier thinkers had not distinguished clearly between 
what has, or is characterized by, a quality and the quality itself. The 
ambiguity of the Greek article-plus-adjective idiom facilitated  
confusion between the two: * the hot' denoted both a hot substance and the 
quality of heat.1 Plato's own indifferent use of 'the beautiful' or 
'beauty' for a Form shows him still haunted by the ghost of this  
misprision, even while thoroughly alive to the distinction between the one 
Form and the many things that 'share in it'. Thus Cornford translates 
the Greek for 'the One itself (???? ?? ??) also as 'unity', and  
represents Plato as drawing the distinction impossible to Parmenides: 'If the 
Real is a whole of parts, it has the property of unity... but it cannot be 
identical with Unity itself (p. 223). This point will concern us again 
a little later. 
(c) Materialists ana7 idealists: the criterion of being B45e~48d). From 
those who try to fix the precise number of real things, the visitor passes 
to 'those who have put the matter in another way'. He depicts a  
continuous Hesiodic battle of gods and giants: the giants will only admit as 
real what has tangible body,2 whereas the gods confine reality to  
'certain intelligible and bodiless Forms' and assign to material objects not 
being but only 'a moving process of becoming'. To make progress it 
must be assumed that some of the materialists (in reality a crude and 
violent lot) have reformed sufficiently to be open to argument. These 
1 Cf. vol. 1, 79 (Anaximander) and 116 (Anaximenes), vol. n, 284 f. (Anaxagoras). Though 
abstract nouns existed, the general term 'quality' (???????, translated by Cicero as 'qualitas') 
was P.'s own invention (Tht. 182a). 
* The obvious reference is to the atomists, for the objection of Wilamowitz (PI. 11, 245) and 
others that single atoms are not perceptible to the senses is beside the point. The atoms are 
corporeal and in bulk tangible and visible, and what is not corporeal (i.e., for the atomists, void) 
is ?? ??. In any case they were materialists in P.'s eyes, which is all that matters (vol. 11, 462). 
I cannot see a strong case for Antisthenes, whom Campbell thought possible (introd. to Soph. 
Ixxiv, though seeming to deny it later, Tht. xl), and Apelt (Beitr. 70 n. 1) indisputable. Some think 
no particular person or school is intended: 'the materialistic tendency in contemporary thought' 
(Campbell), 'the crass unthinking corporealism of the "average man"' (Taylor, PMW 334, 
who for the atomists' view quotes only, without identification, Epicurus), 'die Masse'  
(Wilamowitz). P.'s language suggests that he had a particular school in mind, and my vote goes to 
Democritus. The ' reformed' materialist could be the ordinary man, who would certainly agree 
that 'there is such a thing as justice' (p. 108 n. 4 above). 
138 
Sophist 
will agree that a living creature consists of body and soul, that a soul 
may be just or unjust, and that it is so by its possession, and the  
presence, of justice or the reverse. The soul, they think, is corporeal, but of 
justice, wisdom and the like they cannot deny either their existence or 
their incorporeality. (Reformed characters indeed, if not rather  
puppets, these materialists who meekly acquiesce in the Socratic-Platonic 
language of virtues as entities possessed by, and present in, individuals.1) 
The question Plato is leading up to is this: If corporeality is not 
essential to existence, what criterion of reality can we adopt? Perhaps 
the reformed materialists would agree that whatever has the power or 
potentiality {dynamis) of acting or being affected by action, even to the 
slightest degree, exists, and accept as a definition of the real2 that it is 
nothing but dynamis. On their behalf Theaetetus accepts it * because 
they have nothing better to offer', and the visitor adds that both he and 
they may change their minds later. 
The 'friends of Forms' on the other hand, who distinguish between 
being and becoming, will not admit the new criterion. All power, 
whether active or passive, they relegate to the realm of becoming. Yet 
in their view being can be known and the mind can know it, and if 
knowing is an action, what is known must be affected by that action. 
To be consistent they must deny this. 
This passage raises two related questions: (i) Is the professed  
criterion of reality Plato's own? B) Who are the 'friends of Forms'? 
A) Effectively it existed earlier. Qualities like hot, cold, bitter, salt 
etc. were known by their dynameis—the effects they produced, and 
(more rarely) their potentiality of being affected by others—so that 
dynamis practically became a word for quality, especially in the medical 
writers, at a time when 'the hot' etc. were also thought of as substantive 
1 ?????? and ???????? are too common to need illustration. For ???????? see esp. Lysis 2i7b-i8c 
(cf. Crombie, EPD 11, 255 f.), Euthyd. 301a, Gorg. 497?, Phaedo iood (vol. iv, 278f.); also the 
curious argument at Charm. 1586-59c. I have remarked on these 'substantival expressions' on 
p. 68 above. 
2 Cornford (PTK 238) pointed out that a Spos ('mark') is not necessarily a definition, but the 
words ??????' ????? ?? iv B47d6), taken with e4, cos ????? ??? ???? ?? ???? ???????, seem 
to justify regarding it as intended for a definition here. (Owen also takes it as such in Vlastos's 
Plato 1, 230, n. 14. I find it natural to take ?? ???? as subject of icrnv, against Cornford and 
Runciman, PLE 77 ?. ?.) On the meaning and history of ??????? see Cornford's account (ib. 
234-8), which draws largely on Souilhe^s Etude sur /e terme ???????, summarized in Dies, 
Autour de P. 367-75. 
139 
Parmenides, Theaetetus, Sophist, Politicus 
entities.1 It is reasonable therefore that Plato should propose it ad 
homines, for the acceptance of men with materialistic tendencies, who 
'have nothing better to offer' B47c). A passage in the Phaedrus {170?) 
does suggest that he at one time thought these' powers' at least the most 
important factors in determining a thing's nature {physis): in studying 
a simple object (or the parts of a complex) one must look first for its 
natural dynamis, 'what power it has in relation to what, either to act on 
or be acted on by it'. However, (a) this is not the same as laying down 
this power as a test of the thing's existence; (b) he appeals by name to 
Hippocrates, and the medical writers were especially concerned with 
dynameis? (c) the argument is ad hoc, the subject is oratory, and the 
moral drawn is that to be successful an orator must know what the mind 
can do to what, and by what it can be affected, a thesis that can hardly 
be denied. The phrase 'dynamis of acting and being acted on' occurs 
also at Tht. 156a, where in their theory of sensation the believers that 
all is motion postulate two kinds of it, having respectively these two 
dynameis. As to this, (a) in their account the perceiving sense-organ is 
the passive partner, the sensum the active, whereas in the Sophist the 
knower acts (and the theorists of the Theaetetus equated sensation with 
knowledge) and the known is acted on; (b) the 'all-is-motion' school 
are explicitly distinguished from materialists as a 'cleverer' group {Tht. 
1556-56a); (c) they are in any case not Plato. 
Plato's conversations are separate and artistically composed wholes, 
and appeals from one to another, though often helpful, indeed  
necessary, call for careful attention to the context in each case. Nothing in the 
last paragraph should make us ignore Plato's explicit warning that this 
mark of the real is provisional only ('both we and they may change our 
minds later', i47e), or think of it as anything more than a dialectical 
device to help on the argument,3 his purpose being to bring materialists 
1 See Cornford's account just referred to, and vol. n, 286 n. 1. 
2 With the Phdr. cf. the conjunction of ??????? and ????? in Morb. Sacr. 13 (i, 600 Littr?). 
3 'Only a step, though an important step, in the dialectical progress of the argument'  
(Campbell 124). So also Dies, Bud? ed. 288, but only Apelt, to my knowledge, has made out a detailed 
case for taking it as such (Beitr. 70-7). Aristotle, without mentioning P., cites this definition as 
exhibiting the same flaw as a parallel one in H. Maj.y where, we may note, P. points out the flaw 
himself. (See H. Maj. 297 eff. and Ar. Top. 146821-31. I do not follow Apelt's treatment of the 
latter passage, o.c. 75.) Others adopting the 'dialectical' interpretation of the definition include 
Cornford and Taylor. The view that it is P.'s own is mainly that of an older generation, Grote, 
Zeller, Lutoslawski, Ritter. (Some refT. are in Runciman, PLE 77 n. 2.) 
I40 
Sophist 
and idealists closer together: the former must admit an element of the 
non-material into their world and the latter give up their rigid  
insistence on the immobility and immutability of the completely real. But 
when, a little later, the necessity for motion is to be argued, he adopts 
quite a different approach. 
B) The 'friends of Forms' are those who 'separate Being from 
Becoming and say that we are in touch with Becoming by means of the 
body through sensation, and with real Being by means of the mind 
{psyche) through reason; and that Being is always in the same  
unchanging state, whereas Becoming changes' B48a). These distinctions, both 
ontological (sensibles denied the status of Being) and epistemological 
(sensibles apprehended through bodily organs, Being by the mind alone 
using independent reasoning), agree exactly with the teaching of 
Socrates in the Phaedo and Theaetetus (i84b-86b).1 As for the  
impossibility that Being ('the things that are') should suffer any change, this 
was and remained a pillar of Plato's philosophy from Phaedo to 
Philebus. It is applied to the Forms, and they are repeatedly said to be 
the only realities. The phrase used here,' always in the same unchanging 
state',2 is his favourite description of them, and in the Symposium B11b) 
he says of'the Beautiful itself that it is 'never in any way affected'. In 
the Cratylus B39d) it is only because it remains 'always what it is' that 
it can be the object of knowledge. His language there and elsewhere 
shows that he did not think of being known as being acted on (???????) 
in any way. 
Faced with this (and I have multiplied examples in text and notes to 
bring the point home), I do not see how anyone can doubt that Plato is 
preparing the reader for a modification of his own metaphysics.3 It  
remains to see what form the modification takes. 
1 To be strictly accurate, P. says in Tht. that sensation as well as thought is a function of the 
psyche, though in sensation it must make use of the body and its organs (i84d); but no change of 
doctrine is implied. It is the doctrine of Phaedo 79a-d. 
2 248a, ??? ???? ????? ??????? §????. Cf. Pho, 78 c-d (of ??????? ? ????, earlier ???(?), ??? ???? 
????? ??????? {[???, Rep. 479 a> 5°°c> Tim. 29a. At Phil. 6\ e, 'things that become and perish' 
are contrasted with ?? ???? ????? ??? ??????? ??????. Cf. also Pol. 269 d. The words at Symp. 
211 b are ???? ??????? ?????, as in Soph, the friends of Forms deny them the ??????? ??? 
???????. Pho. 78 c-d and Phil. 58 a show also that the Forms comprise the whole of reality: there 
are no 'things that are' except Forms. 
3 Yet it has been a matter of considerable controversy, a useful, but partial, summary of which 
is given by Jowett's editors, Dialogues in, 322-4. (See also Dies, ed., 292 n, 1.) To supplement 
141 
Parmenides, Theaetetus, Sophist, Politicus 
Motion has a place in the real world B48e~5oe). At 248c the visitor 
bursts out: 'But by Zeus, are we to be easily persuaded that motion, 
life, soul and understanding have no place in the fully real—that it 
neither lives nor thinks, but stays still, august and holy, without wit 
(nous) or movement?' They agree that this would be a strange thing to 
say, and that, since thought is impossible without life and psyche, 
motion and what is moved must be allowed to exist.1 On the other hand, 
{{everything were in motion, intelligence would again be excluded, since 
its operation necessitates unchanging objects.2 Reality therefore,  
philosophically considered, must contain both the unmoved and the moved. 
But here a doubt arises. 'The whole sum of what is', they have 
agreed, must be 'both all that is unmoved and all that is moved' 
B49d3-4). But motion and rest are contraries. To say that they 'are' 
(exist) is not to say of either or both that they move or stay still.' What 
is' must be a third category, embracing both motion and rest, which are 
by reason of their association with Being. The consequence drawn by 
the visitor is that reality ('what is') is 'by its own nature' neither at rest 
nor in motion, but this is baffling, for there is surely no third choice: 
what is not in motion must be at rest, and vice versa. The nature of 
reality is as obscure as that of unreality. 
Plato's language here is bewilderingly loose, even for him, and makes 
it almost impossible to judge when he is talking of a state or attribute 
(or Form) and when of a subject in that state or characterized by the 
attribute, a distinction which he has always expressed in his own way as 
that between being a Form and possessing, sharing in or imitating it. 
this, the friends of Forms have been identified with the Megarians (see esp. Zeller's long argument, 
11.1, 522-5), certain Pythagoreans (Taylor, PMW 385 f., Burnet, T. to P. 280; this goes back to 
Proclus; see Field, P. and Contemps, 227), disciples of P. who had misunderstood him (Campbell 
introd. lxxv, Ritter, Essence 176), and Academic opponents of Eudoxus (Cherniss, ACPA 439 
n. 376). 'Don't knows* include Dies, Jowett (in, 337), Field (P. and C.'s 193f.) and Runciman 
(PLE 76). Some of the above, like Dies, Taylor and Field, definitely exclude P., others think 
him a possibility. Some have tried to find a clue by combining the visitor's remark that he is 
acquainted with these people B48 b) with the fact that he comes from S. Italy. I should not like 
to press this. As an admirer of Parmenides who has yet broken away from him, he is too plainly 
a mouthpiece for P. himself. Those who take a view similar to the one expressed here include 
Grote {PI 11, 458), Friedlander (m, 265), Ross (PTI 107), Grube (P.'s Th. 41, 295 f.), Ritter 
Essence 175), Allan (intr. to Stenzel's PMD, xvii n. 1). 
1 Psyche has been defined in Phdr. B45 c, vol. iv, 419-21) as self-mover and source of all 
motion (kinesis; see p. ??? ?. ? above for the wide range of this word). 
* This was demonstrated, against the neo-Heracliteans, at Tht. i8id-83c. 
142 
Sophist 
The confusion between the abstract nouns 'motion' and * rest' (?????? 
and ??????) and the verbs 'to be moved' or 'at rest' (????????? and 
???????) with their participles, seems complete. (In the previous  
paragraph I have tried to give literal equivalents for Plato's Greek.I At the 
moment Plato does not distinguish (i) '? is neither A nor By meaning 
' ? has not the property (or is not in the state) A or By from (ii) ' ? is not 
identical with A or By; in this case 'what exists is neither at rest nor in 
motion' from' existence is not the same thing as rest or motion'. Having 
said that both moving and unmoved things exist, he might be expected 
to ask, do we mean by this that they are (either one or both classes) 
moving or at rest? And although to say that they exist is not the same 
thing, it does not entail the absurd consequence that neither is either 
moving or at rest, which follows from the (surely unjustified)  
substitution of motion and rest for moving and stationary things. To the end 
Plato continued to believe that Forms were the supreme examples of 
their characters. (Cf. p. 43 above.) 
Either Plato knew what he was doing, and this shifty behaviour2 is 
practised on young Theaetetus (as his Socrates often uses fallacies for 
good endsK to lead him on to the new and important doctrine of the 
combination of Forms, which at least tries to clear this confusion up; or 
he is once again allowing himself a carelessness of expression which in 
this case is hard to forgive. A similar vacillation in the description of 
reality as power, which if it does less harm is at least pointless, makes 
one's doubts difficult to dispel.4 
The problem of motion and reality. Scholarship is sharply divided about 
Plato's meaning in this section. The main questions are: A) Does he 
mean to attribute change to the Forms themselves,5 or simply to enlarge 
1 At 25ob7 Comford calls ?? ?? 'realness', and he may well be right, though I have kept to 
'what is'. Contrariwise ??????? and ???(? at 248aio-n are used (as often) as collectives for 
?? ????????? and ?? ????. 
2 Mentioned but surely played down by Cornford, PTK 248 f. 
3 Cf. pp. 41 f. above. 
4 247d8-e3 * Whatever possesses power really exists' and 248 c 'Whenever power is present to 
something it exists'; but 24867 'Existing things are nothing but power.' 
5 It is true that, as Dies pointed out (ed. 287^), P. occasionally applies ??????? so widely as 
practically to eliminate any dynamic connotation. At Pho. 97 c he speaks of ? ????? ? ???? ?????? 
?????? ? ???????, and similarly at Parm. 136b ?? ????? ??? ??? ????? ??? ?????? ???? ????? 
?????????. If even to be is a ?????, as well as to be known, one must agree with Runciman 
M3 
Parmenides, Theaetetus, Sophist, Politicus 
the realm of Being to include life and intelligence which are not Forms? 
B) Is he going even further in dissent from the friends of Forms and 
admitting what they called Becoming—changing and perishable objects 
of the physical world—as part of the realm of true Being? 
Plato's language makes the second question almost if not quite  
insoluble. Cherniss affirms (SPM 352) that the motion (??????) 
admitted is the Form of motion, and its manifestation in Being is the 
self-motion of life, a non-phenomenal motion which is entirely different 
from ?????? (becoming), what Cornford and de Vogel call * spiritual 
motion'. The disjunction between Being and Becoming is neither  
rejected nor qualified. This is eminently reasonable, and indeed the idea 
that Plato should ever have allowed the sensible world to cross the 
bridge between Becoming and Being is contradicted not only by the 
Republic but by every other dialogue early or late. Apart from the 
Timaeus one has only to look at Philebus 59a-c, where cosmologists 
are said to study 'not the things which always are but those which 
become', whose instability forbids any exact knowledge or thought 
about them. 
Here however he says that Being includes not only the Forms Rest 
and Motion, but also 'whatever is unmoved and whatever is moved', 
which gives colour to the view of Dies, Solmsen, and more recently 
Keyt,1 that it includes all or some things in the physical world. This 
would indeed be a recantation, and it is fair to recall what we have just 
noticed, that Plato can switch from motion-rest to moved-unmoved 
with a seemingly callous indifference to his readers' comfort. Plato, says 
Solmsen, 'would not easily allow himself to consign the Universe in its 
entirety', with all its qualities of order, structure and harmony, 'to 
(PLE 81; cf. 23 n. i) that there is 'a plausible sense in which P. can have thought that the Forms 
can change without forfeiting their changelessness'. Nevertheless since the decision that whatever 
exists is or possesses a ??????? leads directly to the momentous discovery (so at least the visitor 
seems to announce it) of the presence of movement, life and intelligence in the realm of Being, 
the word here must surely have its stronger, more usual sense. In any case we have P.'s earlier 
statement that a Form does not ??????? at all. (See p. 141 n. 2 above.) 
1 Dies, Autour de P. 560 ('dans le visible meme'), Solmsen, P.'s Th. 80-3 ('the Cosmos'), 
Keyt in PQ 1969, esp. p. 6 ('ensouled, living bodies': K. provides further reff. for the discussion). 
See also de Vogel, Philosophia 1, 176-82, 194-209, Ross, PTI 108-11, Grube, P.'s Th. 295-7. 
I have not gone all the way with de Vogel, though her explanation is both attractive and well 
defended, namely that for P. the intelligible world was an articulate, organic unity and therefore 
a 3cpov, the ?????? 3??? of Timaeus 39??. 
144 
Sophist 
"non-being"'. Of course not. He assigns it to Becoming, having taken 
great pains (in the Republic) to show that the choice is not simply 
between the two contraries Being and Non-being. The Timaeus makes 
the position of the cosmos clear: the order which it undoubtedly  
exhibits it owes to its creation by the divine Mind as a copy (?????) of the 
world of eternal Forms B8 b), * fairest of all things that become' B9 a). 
Its status is laid down in the Sophist's sequel, the Politicus:1 
To be always the same and in the same unchanging state belongs only to the 
most divine of all things, and body is not in this class. What we call world and 
cosmos has received many blessed gifts from its creator, but all the same it 
partakes of body. It cannot therefore remain for ever changeless, though its 
motion is as far as possible uniform, invariable and in one place [i.e. circular]. 
The cosmos is alive (???????, 'ensouled', in Greek), and what Plato 
now admits to Being is not the revolving body of the cosmos or its 
contents but the element of soul in it (' life, psyche and understanding', 
248 e), reaching down to subordinate living creatures, and a fortiori the 
supreme Mind, the Creator, who made it' as like as possible to himself 
(Tim. 29a).2 Previously, as we have seen, this status was reserved for 
the Forms alone, and in the Phaedo, though proclaiming the soul 
immortal, Plato ventured only to call it akin to the Forms, resembling 
them, and belonging to the same region G2d-e). From this to parity 
with them as equally belonging to true Being was not a long step—was 
perhaps implicit in the Phaedo—yet the visitor hails almost as a  
revelation his recognition of its full significance in introducing motion and 
activity into what had been a world of static and changeless Forms 
alone. Since the Phaedrus soul has been by definition the self-mover 
that initiates all other motion, and the Laws shows that, though it 
imparts physical motion to bodies by animating them, its own motions 
are spiritual. There the priority of soul to body, and its causal function, 
serve as evidence that psychical activities like wish, reasoning, memory, 
precede corporeal attributes like spatial dimensions and physical 
strengths 
1 269d-e, repeating in substance Rep. 53oa-b. Even more telling is Tim. 38c: the world is 
???????, its model ??. 
2 The middle dialogues make no mention of a supreme Mind, and only once, in passing, of a 
creator {Rep. 530 a). 
3 Phdr. 245 cff., Laws 895e~97b. See vol. iv, 419-21, but also p. 295 n. 3 below. 
11 I45 ghg 
Parmenides, Theaetetus, Sophist, Politicus 
The Forms, as I have said, remain unmoved and impassive. Their 
causal function, in which some have seen a kind of motion,1 resembles 
rather that of Aristotle's Unmoved Mover: the mere existence of their 
perfection excites the development of potentialities in physical things, 
which strive, in the Phaedd% terms, to emulate them. This is perhaps to 
emphasize the 'paradigm' conception of the Forms at the expense of 
the notions of 'sharing' and 'presence in. But of all the expressions 
with which Plato tried to convey the relationship between Forms and 
particulars, that of pattern and copy does seem to go nearest the root 
of the matter. 
Note. I would be the last to claim certainty for what is said here. The 
temptation is always with us to adapt Plato's words to a preconceived idea 
of his philosophy, and readers will be well advised to study the  
interpretations referred to in p. 144 n. 2. Perhaps I might venture a few tentative 
comments on Keyt's. 
(i) p. 4: 'It is only to bodies that the word ??????? can properly be 
applied.' But (p. 5) at 24667 ???? ??????? describes a mortal creature 
(?????? 3???), as at Pol. 261b ?????? (earthy animals) are a class of 
?????????. ?? ???????? ?? is certainly not ??????, and nothing corporeal 
can be ????????, not even the cosmos, which will last for ever only because 
its creator wills it so {Tim. 37C-38C, 4ia-b). Language is not always 
perfectly logical, and I believe that at 249310 P. meant by ?????? simply 
* alive', even with the life of an incorporeal deity whose only activity is 
thought. 
(ii) pp. 7f. K. sees two paradoxes, depending respectively on the statements 
that Forms become, and that they are, known. I do not find this antithesis in 
the text (the present infinitive ???????????? need not mean 'come to be 
known' any more than ??????? need mean 'come to be seen') and we need 
not suppose (though it is possible) that P. has given up his faith that after the 
right intellectual preparation, Forms are known in a single instantaneous 
flash or vision, without any process of becoming known. 
(iii) p. 10. K. claims that P.'s argument at Crat. 4396-403 does not prove 
that an object of knowledge must be completely changeless, but only that it 
must not be always changing. Proteus is changeable, but for so long as he 
chooses to remain, say, a tree or a leopard, I can, on K.'s argument,' know 
this'. But (a) According to this reply, it is only so long as he is changeless 
1 For these and other theories, e.g. the 'animation' of the Forms as conscious and thinking 
beings, reff. will be found in Dies, Soph. 288 n. 1. 
I46 
Sophist 
that he is knowable: an object is still only knowable in so far as it is  
changeless; (b) K. adds, * In my example I am only interested in knowing what he is 
now, not what he is really', which surely abolishes any relevance to P., for 
whom to know something was to know its essence, 'what it is really', and 
nothing else. 
The interrelationship of Forms B5oe~54b). A little recapitulation may 
help. We call the Sophist a producer of counterfeits and falsehoods, 
unrealities in fact. He takes refuge in Parmenides's dictum that there is 
no middle way between 'what is' and 'what in no way is', 'non-being 
in itself'.1 To catch him we have to show that Parmenides was wrong, 
that 'what is not in some respect is and conversely what is in a way is 
not' B40e, 241 d). We started with an examination of past theories, of 
cosmologists who claimed that only very few basic constituents of the 
world exist and everything else has only a derivative status, becoming 
and perishing as the substantive elements combine and separate, and of 
Parmenides himself who allowed only one Being. Then, from another 
angle, we criticized the warring factions of materialists and idealists for 
their extremism. Reality, what ?, must include both the unmoved and 
unchanging (the Forms of the idealists) and motion, at least in the form 
of life and intelligence. But what does it mean to say that Motion and 
Rest both are} Not that they are the same (for they are contraries), nor 
that either or both is the same as Being. Being is a third thing, but this 
cannot mean that it is neither in motion nor at rest, for that is nonsense. 
What is needed is a thorough investigation of the possible meanings of 
'is' and 'is not'. So far, says the visitor, both have baffled us. As we 
proceed, one may throw light on the other, or if both elude us, we may 
hope to steer a course between them.2 
So with characteristic skill Plato has taken us in living discussion 
from the Sophist's wiles, by way of early cosmology, monism,  
materialism, and idealism, to an analysis of the concepts expressed  
indiscriminately by the Greek word for 'to be' as ordinarily used.3 These 
were in the main three: identity (Tom is my son), attribution (Tom is 
1 238d9 ?? ?? ?? ???' ????, 240?2 ?? ??????? ????. 
2 25oe~5ia, adopting Campbell's and Owen's rendering {Plato 1, ed. Vlastos, 230) in  
preference to Cornford's. 
3 243 d. Tht. 'You mean we must first enquire what people who speak of" what is" think they 
are signifying.' Vis. * You have taken my meaning exactly.' 
I47 11-2 
Parmenides, Theaetetus, Sophist, Politicus 
tall), and existence (poor Tom is no more).1 It has been commonly 
thought that in the Sophist Plato recognized all three. Cornford saw 
only two, existence and identity (PTK 296) and it has recently been 
claimed that he did not distinguish the existential sense from either of 
the other two.2 What is certain is that the trouble started from Par- 
menides's assertion that 'is not' could only apply to the absolutely  
nonexistent, 'what is not in any way at alF, sheer nonentity. 
Plato's new approach to the study of what is meant by being and not 
being starts from the age-old question, brought to a head by Par- 
menides, of the One and the Many. We habitually call a thing by many 
names, as when we say a man is pale, tall, good or bad; but, say some, 
you cannot make one thing many. Since man and good are different, 
you cannot legitimately say a man is good, but only 'man is man' and 
1 I do not think this classification will mislead us at present, though Kahn sees the primary 
sense of ????? as 'to be the case' or 'true' (an important sense to which I have referred more 
than once: see p. 69 above), with both existential and predicative senses as special cases of this. 
He maintains indeed that the Greek notion of being differs radically from ours of existence. For 
present purposes at least I regard usages like Homer's and Hesiod's ???? ???? ?????? and the 
famous ?*???? ??????? (Aesch. Ag. 958) as both existential and primary, but Kahn's article ('The 
Greek Verb "To Be" and the Concept of Being', Foundations oj Language 1966) should be 
required reading for any student of Greek thought. 
2 By Malcolm in Phron. 1967, 130??., Owen, Plato 1, 223 ff., and most recently Gosling, Plato 
A973), 213 ff. M. cites also Runciman, PLE 84, but R.'s position, if one reads on to p. 90, is a 
little delicate. 'P. does not consider the problem of existence as such. He establishes only that 
everything must have some sort of Being; but this cannot be said to be the same thing.' 'In two 
places the purely existential sense does appear to be what is meant... P. is deliberately [my italics] 
using ????? in a sense where it is legitimate and complete without the addition of anything to 
convert it into an identitative or copulative use.' Yet 'he still did not specifically distinguish the 
existential sense as such'. On p. 102 R. calls it the principal achievement of Plato's analysis that 
'it shows how negation need not involve an assertion of non-existence'. 1 am not sure how this 
could be achieved otherwise than by drawing the distinction between the copulative and 
existential senses of ?????. Bluck, I now see, held that 'Plato assimilated to each other the 
existential and the copulative senses of "to be".' See his Soph., published 1975, 62-7, 119. At the 
risk of over-simplification (certainly not a fault of P.'s recent interpreters) I suggest that when P. 
asserts that Motion and Rest both are (l^d 10), but are not identical, he shows himself aware of 
the distinction between the ' is' of existence and of identity (' identity and existence cannot be the 
same thing' is a fair translation of 255 C3); and by introducing the asymmetrical relationship of 
one Form to another (p. 117 above: ???????? is the verb commonly used of the relation of 
particulars to Forms) he draws attention to the third use of'to be', the attributive, predicative or 
copulative. (I have been encouraged here by the lucid and convincing article of Ackrill in SPM 
207-18.) 
R. Robinson is of course right to remind us that for P. all this was not grammar or logic but 
ontology {Essays 37). 'He is talking about Being, not the word "being".' But this need not 
convict of error those who (like Shorey and Taylor) have ascribed to him the discovery of the 
copula or claimed that he distinguishes meanings of 'is' and 'is not'. Cf. also R.'s next para.: 
P. 'gives us an account of what he calls the " form" of the Other; there is no such form;  
nevertheless, all that he says about it is true of something else, namely the word "other"'. 
I48 
Sophist 
'good is good'. This however is quickly dismissed as Entertainment 
for boys and stupid old men'.1 It had of course been settled by the 
doctrine that one individual could partake of, or be associated with, 
many Forms.2 The philosophical question (as it is called later) is 
whether Forms themselves can associate or combine with each other, 
a question foreshadowed by the Parmenides, where Socrates said it 
would at least be a marvel if contrary Forms could combine, Similarity 
with Dissimilarity, Plurality with Unity, Rest with Motion and the 
like.3 Now the doctrine of Forms enters a new stage with a full  
consideration of all possibilities: no Forms can combine, all Forms can 
combine, some can combine with some others, some can combine with 
all others. To disentangle their various relationships is the subject of a 
special science, dialectic.4 But good Heavens! (says the visitor), in  
pursuing the Sophist we seem to have stumbled first on the philosopher, 
whose province it is. Well, he must wait his turn.5 Just now the Sophist 
remains the quarry. His hiding-place is in the darkness of not-being, 
and to find him necessitates going into the question for at least some of 
the Forms. 
1 The argument against all except identical predication is generally attributed to Antisthenes, 
but I have expressed doubts about this in vol. m B14, 216-18). 
* (That the arguments of Parm. do not imply abandonment of the earlier theory of Forms I 
have ventured to maintain on pp. 58-61 above.) Note the wording of 252bc>-io. The opponents 
of non-tautological predication 'do not allow anything, by partaking in another property 
(???????? ?????? ?????????), to be called that other'. Though this makes for awkward English, 
I doubt if the genitive ?????? depends on ????????? (Cornford, Owen in Plato 1, 251 n. 48, 256), 
which would be decidedly awkward Greek. A man may not share in a property (goodness) which 
is other than himself (man) and so be called 'good' as well as 'man'. 
3 Similarly at Pho. i02d-e, S. says that Largeness will never 'admit' Smallness. For ???????? 
cf. Soph. 253ci. The various terms used to describe the relationship between Forms in Soph. 
(?????????, ???????? etc.: full list in Cornford, PTK 255) are mostly those which P. regularly 
employs for that of particular to Form. Now at least there is no doubt that Motion, Rest and so 
on are Forms, not moving etc. things, though we have to wait for a general term until 235b8 
(????), d5 (????) and 254c2 (????). 4 For its full description at 253d-e see p. 129^ above. 
5 Peck {CQ 1952, 45) took this as a warning that the arguments which follow will be sophistic, 
not philosophic, thus supporting his general thesis that the ???? or ???? in this dialogue have 
nothing to do with Platonic Forms. But it does not appear that P. first puts forward sophistic 
arguments and then corrects them. The ???? belong to the main line of reasoning by which the 
Sophist is finally run to earth, and e.g. the proof that ?????? and ??????? are ???? in their own 
right (Peck 46 f.) does not commit the fallacy of Dionysodorus. The dropping of the predicate in 
B) (see ib.) is immediately put right by D). To rely on unsound argument to this extent would 
be no way to define a Sophist, a philosopher or anything else, nor is P. hinting that he will do so. 
Cherniss in JHS 1957 A), 23 n. 57, points out that ?????, ?????? and ??????? appear as Forms in 
Tim. C5 a, 37a-b), which Peck thought later than Sophist 254b 3 is another pointer to the intended 
dialogue on the philosopher. 
149 
Parmenides, Theaetetus, Sophist, Politicus 
In later terminology, that a Form {A) ' shares in' another (B) means 
that it has ? as an attribute, so that ? can be truly predicated of it, just 
as an individual, Socrates, can share in wisdom and so be wise. The 
relationship may be reciprocal, in which case the verbs Combine' or 
'associate' (??????????, ?????????) are apt, or it may not. Motion shares 
in (??????? ???) Being (for there is such a thing), but not Being in 
Motion, which would mean that whatever is (exists) moves.1 
It cannot be either that no Forms can combine, or that all can  
combine with each other. The former would mean that no Form could even 
exist (that is, in Form-language, partake of Existence). Later he says 
that it would abolish all discourse, which consists in weaving Forms 
together. But if all Forms could combine, even self-contradictory  
statements would be true, such as that Motion is at rest or Rest in motion 
B52d). 
It might well be said that the Form Motion must be unmoving 
( = unchanging), for all Forms are. I do not myself see that Plato ever 
completely overcame this difficulty, which is bound up with his  
indiscriminate use of abstract noun ('Motion', ???????, or 'Being', ?????) 
and participle ('the moved', ?? ??????????, or ' what is', ?? ??). Forms 
impart their qualities to whatever shares in or combines with them by 
virtue of being themselves the supreme and perfect examples of those 
qualities. It is curious how this hiatus in Plato's thought is ignored by 
Cornford, Ross (see ???112 {.) and others. Plato may possibly have 
had in mind that contrary Forms cannot mingle in (enter simultaneously 
into) the same individual, as the Phaedo explained (io2d-e); but that 
is not what he says. The doctrine of Forms independent of their 
instantiations had originally a strong metaphysical or religious flavour 
(they are 'divine') necessitating liberal use of metaphor, and as purely 
logical tools they play an uncongenial role—those Forms of which the 
Platonic Socrates said 'in his simple, unsophisticated way' that by some 
sort of attachment to things (and he could not safely say what sort) they 
gave them the character they had (Pho. iood). 
1 P. also marks the difference by using koivcoveiv and its compounds with genitive or dative. 
See Ross, PTI111 n. 6. 
150 
Sophist 
Five of the greatest Forms: Parmenides refuted B54b~59b). The only 
remaining possibility is that some Forms can, and some cannot,  
combine. In this they resemble the letters of the alphabet, and perhaps also 
in that some of these (the vowels) blend with all the others and make 
possible their union with each other B53 a, 254b-c). Such all-pervading 
or penetrating Forms would be either bonds enabling others to combine 
or conversely in some cases responsible for keeping them apart.1 The 
next step is to examine the relations between actual Forms in the light of 
these generalities, and since to review them all would be impracticable, 
the visitor picks out 'some of those recognized as the greatest' B54 c),2 
which are especially relevant to a clear understanding of'what is' and 
'what is not'. Three are already familiar: Motion, Rest and Being, of 
which the first two will not mix, but the third mixes with both, for both 
are. Further, each is the same as itself and different from the others, and 
since none is identical with the Same (the Form Sameness), Same and 
Different are two more Forms, in which the first three share.3 In  
demonstrating that Being and Difference are not identical, Plato draws the 
distinction between a thing's being ' in and by itself and being relative 
to something else. The Form Being includes both, i.e. Socrates is 
(exists, or is himselfL and Socrates is . . . (e.g. shorter than Simmias) 
but Difference is always relative (???$ ??). 
Attention is now concentrated on Difference, because the Sophist 
relied for his escape on the impossibility of saying of anything that it 
'is not', and Plato wants to show that it may be equivalent only to the 
perfectly permissible statement that it' is different' from something else. 
1 253b-c. See Cornford, PTK 261 f. 
* For the translation 'greatest' rather than 'very great' (Cornford) see Ross, PTI 113 n. 6. 
Peck (CQ 1952, 45) says 'the meaning of ???????? is not explained'. Presumably it means, as 
usual, 'largest', i.e. widest. Each of the ???? Being, Same and Other includes everything, and 
Rest and Motion divide the whole field between them. Why should this need explanation? 
Cf. also Ackrill in Ryle, 391. Others (Leisegang, RE 2495, Trevaskis in Phron. 1962) render 
it 'most important' or 'basic', because these are the ???? to which earlier philosophers had paid 
most attention, asking 'What is Being?' and answering 'It is motion' (Heraclitus) or 'rest' 
(Parmenides). 
3 That 'Motion is different (from Rest)' does not mean that Motion and Difference are the 
same thing seems obvious enough, but P. defends it by quite a complicated little argument, 
explained by Cornford, PTK 280 n. 1. 
4 Owen has suggested that the contrast is probably not between the complete and incomplete 
uses of ????? but between two incomplete uses, in statements of identity and of predication {New 
Essays, ed. Bambrough, 71 n. 1). It may, however, distinguish the absolute use from use as a 
copula. 
151 
Parmenides, Theaetetus, Sophist, Politicus 
Difference is a 'vowel' Form which pervades' all the others B55c), 
for each is different from the rest without being Difference itself. We 
are now getting to the heart of the matter. We can say Motion is both 
the same and not the same because we are not using the word in the same 
sense in both cases (??.. .?????? ????????? 256a11). Our sentence is in 
fact incomplete. Properly speaking it should run: It is the same  
(partakes of Sameness) with reference to itself by being the same as itself, 
but not the same in that it partakes in, or combines with, Difference in 
relation to everything else. Thus the principle that whereas a particular 
could possess contrary qualities by partaking in contrary Forms, this 
could scarcely be true of the relationships of Forms to each other (Parm. 
I29d-e, p. 37 above), is modified. A Form cannot mix with its own 
contrary, but any Forms, say A and 2?, can partake of others X and K, 
X and ? being contraries, if the concepts of which X and ? are  
prototypes are relative and with the proviso that A and ? cannot partake of 
both in relation to the same thing. Completed and philosophically 
analysed, sentences like 'A is the same' and * A is different' become ? 
partakes of Sameness with reference to itself and ? partakes of 
Difference with reference to B.' And it both ' is' (partakes of Being) 
and Ms not' (is different from Being itself). Even Being Ms not' in the 
sense that it is different from the other Forms B57a). 
Perhaps the great contribution of the Sophist to philosophy lies in 
the statement that I have italicized, that a word can be used in more than 
one sense.1 The whole challenge of Parmenides, and many of the  
arguments of Sophistic, rested on the assumption that the verb ' to be' meant 
one thing and one thing only. Once it had been shown that the same 
word was not always used to express the same concept—that for 
instance existence, identity and attribution we*e not the same, though 
expressed by the same word Ms'—Greek thought was freed from a 
whole host of unreal (and at this distance of time almost  
incomprehensible) problems. Aristotle could start where Plato left off, with the 
simple pronouncement that ' there are many ways in which a thing is 
said to "be"', and proceed without further ado to enumerate them, 
1 Cf. also 259c-d, where the visitor speaks of the importance, 'when anyone says that  
something different is somehow the same, of being able to determine in what sense and in what 
respect he means it is one or the other'. He adds that this, in contrast to the trivialities of eristic 
argument, is a task both difficult and worth while. 
152 
Sophist 
discarding the clumsy language of Forms which had served its purpose 
in the pioneer work of Plato.1 
As with knowledge B57 c) and its different departments, sciences or 
crafts,2 'the nature of the Different' is divided into parts', the not- 
beautiful, not-tall, not-just and so on. These are contrasted with their 
opposites the beautiful and so on, and exist no less than they. 
The not-beautiful obviously includes everything that is not beautiful, 
the not-just everything that is not (is different from) just (thus  
disposing of the fallacy committed by Socrates at Prou 331a: the difference 
between contraries and contradictories has just been cleared up at 257b). 
Whether * everything' should be ' every Form' the fatal ambiguity of 
'the beautiful' makes it difficult to decide.3 Probably the not-beautiful 
is, as Cornford said (PTK 293), 'the collective name for all the Forms 
there are, other than the single Form "Beautiful"'. Each of them  
partakes of the Different with reference to the others. This has a bearing 
on the warning in the Politicus to divide according to true kinds or 
classes. To divide mankind into Greeks and barbarians is bad  
classification because' barbarian' is simply a name covering all the heterogeneous 
races which are not Greek—Lydians, Phrygians, Persians, Egyptians 
and so on. It signifies a part or portion (?????) of humanity but not a 
true species (?????) with its own character. In the language of the 
Sophist it applies to all races which participate in Difference with 
reference to Greeks.4 
1 ?? ?? ??????? ???????? are the opening words of Metaph. Z, and in his glossary of  
philosophical terms in bk. ? he can begin each section by saying that 'cause', 'nature', 'quality', 
'state', or whatever it may be, is in one sense (lit. 'way of speaking') *, in anothery and so on. 
* Cf. Rep. 438c-d: 'Knowledge as such is of subject as such, but a particular science is of a 
particular subject, e.g. there is a knowledge of building, set apart from the others by the name 
architecture, because it has a character different from the others.' 
3 It is not always easy to share Cornford's optimistic view (p. 292) that though P. uses ?? ?? 
etc. ambiguously (sometimes as 'Existence itself, sometimes as 'the existent' or 'that which is 
so and so'), he himself was always aware of the ambiguities. Was Cornford himself always clear? 
In his translation of 258 c ('that which is not ...is what-is-not, a single Form') should not either 
?? ?? ?? be 'Non-being' or alternatively lv ????? 'a single class', with no overtones of capital- 
letter Forms? 
4 Pol. 262c-63b, pp. 168, 293 n. 3 below. As an illustration P. also cites Number, of which 
subordinate Forms (Odd and Even) are species or parts (????: every subordinate ????? is a 
?????, though not every ????? is an ?????, 263 b). Schipper's statement that a Form has no parts 
(Phron. 1964, 43) is erroneous. The language of the early dialogues (the Pious a part of the Just, 
Courage a part of Virtue, Euthyphro i2d, Lack. 199 ?) was not abandoned when the doctrine of 
Forms developed further. Cf. p. 276 below. 
153 
Parmenides, Theaetetus, Sophist, Politicus 
It looks as if the chase is over. By noting the all-pervading nature of 
Difference Plato has been able to maintain against Parmenides that what 
is not really is, though contrasted with what also is, as the non-Beautiful 
is contrasted with the Beautiful but exists no less than it B58d-e). 
Non-being need not mean absolute non-existence, but simply Difference 
in relation to x, so that' there is much that each Form is, but an infinite 
number of things that it is not' B56c).' And this is the non-being which 
our concern with the Sophist led us to seek' B58b6). When we accuse 
him of creating misleading copies that 'are not' the originals, we only 
mean that they differ from them. 
Speech and thought: the nature of falsehoodB5^-264^. But the visitor 
is not yet satisfied. The Sophist, it has been said, deceives us, that is, 
makes us believe falsehoods B40 d), but thinking what is not true is 
thinking what is not in yet another sense more difficult to explain.1 The 
difficulty here arises from the Greek equation of truth with 'being', 
referred to several times already. A Greek spoke not only of saying or 
believing a falsehood (pseudos) but equally idiomatically of saying or 
believing what is not.2 So far the argument has not accounted for 
falsehoods as satisfactorily as for negative statements. 
The Sophist had two escape-routes. First, he denied that he could be 
a deceiver or illusionist because it is logically impossible to speak or 
conceive of what is not. This is the difficulty stated and left unsolved 
in the Theaetetus, and the reason given for the failure was that they were 
wrong to discuss false judgement before deciding the prior question of 
the nature of knowledge. It is at least possible that the Theaetetus as 
well as the three following dialogues formed together a single plan in 
Plato's mind. The stages would be: A) Preliminary: to raise the  
question of knowledge, explore its difficulties, and reject some inadequate 
answers {Theaetetus); B) Destructive: to expose as fraudulent the 
Sophists' claim to knowledge {Sophist); C) Positive: to answer the 
question and describe the state of mind of the possessor of genuine 
knowledge {Politicus leading up to Philosopher). The Politicus fills some 
1 The impossibility of false statement or judgement has been amply illustrated in the Euthyd. 
and is also asserted by Cratylus at Crat. 429d (p. 12 above). That its object cannot be 'what is 
not' was affirmed in Tht., where the problem here tackled was shelved (p. 107 above). 
2 240d ?????? ??? ?? ?? ???? ????^??? ??? ????? ?????, Cf. p. 69 above. 
154 
Sophist 
gaps in the right method of acquiring knowledge and describes the true 
statesman, who is in fact the philosopher in one of his aspects, and the 
series would have been crowned by a description of philosophy and 
the philosopher as such. 
Now however it has been shown that 'is not' may mean only 'is 
different from \ This the Sophist must accept, but he may still hold out 
by suggesting that, since it also appeared that not all Forms can  
combine, Speech and Belief may be of those that cannot combine with Not- 
being, which again would make falsehood and deception impossible. 
To counter this new wile will necessitate an investigation into the whole 
nature and status of logos (here significant speech, statement or  
continuous discourse), doxa and phantasia. Doxa (belief or judgement) is 
the outcome of thought, which is simply a logos carried on silently by 
the mind with itself B636-643; Tht. i89e-c)oa), and when it depends 
on sensation is called phantasia. Logos, therefore, is basic to all three, 
and a study of it will be rewarding for its own sake, since without logos 
there could be no philosophy. 
In this connexion the visitor says that those who deny any  
combination of Forms annihilate all logos, ' for the Logos owes its birth to the 
weaving together of Forms with each other'.1 If there were no blending 
of Forms, Logos could never combine with Being, i.e. could not exist.2 
Having removed that initial difficulty, we have now to consider its 
nature, in the hope of discovering whether or not it will blend with Not- 
being, so allowing for 'a logos of what is not', or falsehood. 
A logos consists of words, which are of two sorts, nouns and verbs.3 
Like the realities which they express, some words can combine and 
others not. The simplest logos must 'weave together' B62d4) one of 
1 The context shows that at 259? the 'things' separated (no noun is used) are in Plato's eyes 
Forms. The reference is to the foolish people of 251 b who delight in allowing none but identical 
predication, and the question at 251 d: 'Are we not to attach . . . any Form to any other, but to 
treat them all as incapable of mingling or partaking in one another?' 
3 See additional note on p. 161 below, and cf. 260a: 'See how pertinently we opposed such 
men [sc. those who separate everything from everything else] and forced them to allow one thing 
to mingle with another.' 'Pertinently to what?' 'To the thesis that the Logos was one of the 
kinds (????) of things that are.' 
3 ??????? (lit. names, cf. nomen, 'noun') is first used generally for 'word' B61 d2), and 
subsequently B62a 1) confined to nouns, words being defined as 'vocal signs concerned with 
being' (?????, 261 e 5). On the meaning of ???? see p. 11 n. 4 above and cf. Luce, CQ 1969, 229 
n. 1. Here it is confined to ?? ??! ???? ???????? ?? ?????? B6233). 
?55 
Parmenides, Theaetetus, Sophist, Politicus 
each class, for only so can one 'say something' (legein, i.e. make a 
significant statement) rather than merely naming.1 Being about  
something, it will be either true or false; which it is depends on the relation 
between its parts. Once again the idea, already established, that 'to be' 
expresses relation as well as existence, provides the key. The conditions 
for a false statement like 'Theaetetus is flying' are two: (i) that the 
subject must be real (Theaetetus, not Mr Gradgrind),2 B) that the  
predicate must also be real (flying occurs), but not applicable to the subject. 
As a negative statement like ' Socrates is not a Spartan' or' Clinias is not 
wise' expresses not non-existence but difference, so the false statement 
about Theaetetus expresses not 'what is not' absolutely but 'what is 
other than' the things that are concerning Theaetetus. The true one 
states 'the things that are about him as they are' B63b-d).3 
The extent of Plato's achievement here is well summed up by Ackrill 
(Monist 1966, 383 f.). He 
makes clear the special grammatical and logical complexity of the sentence, a 
unit which interweaves a naming part (?????) with a saying part ((?>???). A 
statement must be about something, and it must say something about it. But 
it can assert a real property of a real subject when that property does not in 
fact belong to that subject. 
It is obvious that Plato's brief discussion in the Sophist does not say all 
that needs to be said about false statement. For example his account does not 
cater for the possibility of false existential statements. Nor does he make 
clear in what sense each part of a sentence must stand for something 'real'. 
Nevertheless he certainly makes an important advance in the Sophist by 
recognizing the special type of complexity which a sentence enjoys, and by 
tying the notions of truth and falsity to these specially complex units. 
Return to dichotomy: the Sophist finally captured B660-268 d). The 
visitor now sums up the position. The aim was to define the essential 
nature of the Sophist by the dialectical method of diairesis, but when 
1 On the question whether this is consistent with the treatment of names in Crat., see Luce 229. 
In general, their treatment there should be kept in mind. Cf. pp. 19-23, 25-9 above. That a 
????? consists of ????? and ???? is repeated both there D25 a) and in Ep. 7 C42b). 
2 The status of statements about fictional characters and mythical creatures is still under 
discussion. Strawson favours the view that they are neither true nor false (lntrod. to Log. Theory 
69), others that they are true and others again (including Russell) that they are false. See H. G. 
Blocker, 'The Truth about Fictional Entities', in PQ 1974. 
3 Cf. Ctesippus at Euthyd. 284c, vol. iv, 271. There the Sophists were allowed to triumph: 
here their crudity is revealed. 
156 
Sophist 
they had reached the point of seeing him as 'image-maker', offering the 
appearance of knowledge without the reality (pp. 133 f. above), they 
were diverted by the need to analyse and justify the concepts of  
imitation and deception. This done, they can return to their dichotomies. 
However, instead of continuing from the division of images into  
likenesses and semblances at which they broke off, the visitor goes right 
back to the very widest genus, art, under which the Sophist's  
occupation, like the angler's, could be brought (p. 124), and from which  
therefore the classification had to begin. This is reasonable, because imitation 
or image-making had not been reached by the strict method of  
dichotomy. That had been used to elicit each of seven varieties of Sophistry, 
and its essential nature as imitation (' what a Sophist really is', 231 c) had 
been discovered in the more familiar Socratic way of achieving a 
definition by abstracting the common element from a number of diverse 
instances.1 It is now to be given its place in a complete series of  
dichotomies. 
As image-maker the Sophist belongs to the other main subdivision of 
the arts, productive not acquisitive, though admittedly the acquisitive 
branch did show up certain genuine aspects of him B65 a). To pursue 
the dichotomies, his art is productive (not acquisitive), human (not 
divine), of imitations (not originals), which are semblances (not  
likenesses: see p. 134 for the distinction), mimetic, i.e. in his own person 
(not produced with tools),2 in ignorance (not with knowledge),  
insincere (not naive), in personal encounters, using brief arguments to 
make an adversary contradict himself (not using long speeches). Here 
at last we have the Sophist, rightly called by a derivative of sophos to 
indicate that, though not himself wise, he is an imitator of the wise man. 
So the dialogue ends. 
The last dichotomy repeats one in an earlier diairesis B25 b-c). Plato 
is evidently considering the Sophist purely as an eristic, disregarding 
his epideictic displays. At 223 b, where he is seen as a ' hunter of men 
1 It may also be described as the collection that precedes division (p. 131). The above  
considerations make it unnecessary to suppose with Cornford that Plato's is 'consciously shelving 
the eidola problem' because he cannot yet solve it ('If he had thought it was already solved, he 
would have taken up the Division of Image-making at the point where it was dropped', PTK 
323), and the ontological status of sensibles in Rep. 5 seems irrelevant. As C. himself says, 'the 
only eidola we are now concerned with are those which the Sophist is accused of creating'. 
2 The distinction is between acting a part and representation in sculpture and painting. 
157 
Parmenides, Theaetetus, Sophist, Politicus 
privately', Cornford (PTK 174) supposed that his displays before an 
invited audience are included in contrast to forensic and political  
oratory. But (apart from the fact that the Sophists gave them at the Panhel- 
lenic festivals) they can hardly be included in the ' private controversy 
chopped up into questions and answers' of 225b, where Sophistic is 
classified under agonistic. That only referred to one aspect of Sophistic, 
but now we meet the same thing in its universal definition. This is 
curious when one remembers how the Platonic Socrates complains of 
the way that Protagoras will not stand up to short questions and 
answers but prefers to launch out into long speeches, or conversely 
Hippias is unhappy because Socrates will not allow him to explain  
himself at length instead of replying briefly to questions. Here long speeches 
are assigned to the demagogue, expressly distinguished from the Sophist 
with his contradictious brachylogy. The type seems to be that of 
Euthydemus and Dionysodorus rather than Protagoras. Plato's reasons 
for limiting Sophistic in this way are a matter of guesswork,1 but with 
the 'Sophistry of noble extraction' in mind (pp. i28f. above), we 
may suspect that he still felt the need to distinguish Socrates from the 
kind of Sophist with whom he was most easily confused. 
If in some things the Sophist seems to reveal a changed Plato, the 
final diairesis reminds us that in at least two fundamental points he did 
not change. First, in clarifying the division of imitation into  
knowledgeable and ignorant, and of the latter into naive and insincere, he 
shows himself still mindful of the Socratic search for moral truth, for 
he takes as his illustration those who appear to others to exhibit in  
themselves 'the shape of Justice and the whole of Virtue', deeming to be, 
but in no way being, just', because they have no knowledge of what 
Justice is, but only doxa. Some are simple souls who genuinely mistake 
their doxa for the truth, but others are experienced debaters with a 
shrewd suspicion that what they give out for knowledge is not  
knowledge at all B07c-268a). These of course include the Sophists, and the 
passage would be quite at home in the Gorgias or Meno. 
1 Campbell (ed., p. xlviii) says P. is describing the ideal Sophist rather than any individual. 
Ritter comments simply that the eristic aspect is singled out as the most dangerous: compared 
with it mere loquacity is harmless (N. Unters. 65). In Taylor's view (PMW 376, following 
Schleiermacher as do many others) the solution was obvious: for 'Sophists' read 'Megarians'. 
I would not go bail for this, even though the latter are credited with maintaining ?? ?? §v ????? 
??? ?? §????? ?? ????? (vol. in, 500)· 
158 
Sophist 
Secondly we have the division of production or creation into human 
and divine. To justify it the visitor is at pains to insist that what are 
called the works of Nature must not be looked on as the product of 
some automatic and mindless force, but as the work of a craftsman-god, 
acting with reason, art and knowledge B65 c). The Phaedo had upheld 
a ideological view of the universe based on a development of Anaxago- 
ras's brilliant but unexploited pronouncement that all things were 
ordered by Mind, and the Timaeus, whether written before or after the 
Sophist, expounds in detail the creation of the living cosmos by a divine 
craftsman1 after the pattern of the Forms. In traditional Greek religion 
none of the gods, not even Zeus, created the world, and the  
combination of supreme god with creator may fairly be credited to Plato among 
philosophers, even if he had learned something from the Orphic 
writers to whom he frequently shows himself indebted.2 It is interesting 
to find it in a predominantly critical and analytical work like the Sophist. 
The Sophist and the Forms. The view adopted here of the role of the 
Forms in this dialogue should have emerged plainly enough by now, 
but it is a question that has caused considerable bewilderment. Richard 
Robinson for instance speaks sadly of' the dreadful question of whether 
the ??????? ???? of the Sophist are Forms, on which I have not yet 
succeeded in reaching a confident opinion', and Peck denied that they 
were. Terminology does not help: ????? and ????? (used  
synonymously, e.g. at 254b-c) are frequently given by Plato their common 
meanings of Character' or ' class'. We have also had to face the helpless 
feeling induced by the indiscriminate use of abstract noun and adjective 
(or participle) with the article (pp. 142 f., 150 above). 
Where Forms appear with capital initial in this chapter, it is assumed 
that Plato thought of them as Platonic Forms, not merely concepts of 
the mind but realities with an objective and independent existence. He 
even remembers their exalted status, for at 254a he says that the  
philosopher, through his devotion to the Form of Being, dwells in the 
brightness of the divine, and just before this the philosopher's ability to 
1 ??????????. Cf. ?????????????? at Soph. 265 b4 and ?????????? at Pol. 273 b and of the 
creator of the stars at Rep. 52961. 
2 For Zeus as creator in Orphic literature see Guthrie, OGR io6f. 
J59 
Parmenides, Theaetetus, Sophist, Politicus 
distinguish eide aright, and the ways in which they can and cannot 
combine, is called dialectic, and though the description of dialectic as 
discrimination between eide goes back to Socrates (p. 27 above), its 
goal since the central books of the Republic can only be Forms. A 
moment later he as good as tells us that this is matter for the Philosopher 
B54b). 
Yet this aspect of the Forms merits, and finds, no place in the logical 
problems of the Sophist, and their solution is none the better for  
metaphysical props. Indeed the paradigmatic character of the Forms,  
involving the belief that they are perfect instances of themselves, only 
causes trouble. At 255310 Plato meets an argument by rejecting as 
absurd the idea that Motion can be at rest. This ought to show either 
that Motion is not a Form or that motion and change have been not 
only accepted as realities but introduced into the Forms themselves. 
Yet it is hard to believe that Plato intended to imply either conclusion. 
The metaphorical language inseparable from the doctrine of Forms— 
sharing, binding, running through and so on—and especially its undue 
reliance on substantival expressions,1 were unsuited to exploring the 
fields of logic and language. Difference is for Plato not simply a relation 
but a relational property (and so a Form), which leads to the clumsiness 
of analysing' Motion is not Rest' into ' Motion shares in Difference with 
respect to Rest', instead of the simpler differs from'. It is said that he 
was hampered by the resources of language available to him at that 
time, but there was no difficulty about saying notion differs from rest' 
in contemporary Greek. What hampered him was his faith in the all- 
sufficiency of the doctrine of Forms, and the undoubted logical  
advances of the Sophist were gained in spite of that faith, not through it. 
By his definition of a logos he confined propositions to the subject- 
predicate type (the predicate at least being for him a Form), which 
persisted in Aristotle {Rhet. 1404 b 26) and dogged the footsteps of 
logic until the twentieth century. Like everyone else, he could not 
escape completely from his historical situation. Parmenides had  
compelled him to argue at length a case for 'is' not always meaning exists, 
and faced with this necessity he thought of the Forms, which he had 
first evolved in response to the Socratic faith in absolute values, as 
1 Commented on in vol. iv, 226. 
160 
Sophist 
suitable instruments for the purpose. They do play a role in the Sophist, 
but philosophy might have progressed more easily if they had not. 
ADDITIONAL NOTES 
(/) On 259d {p. 155)/ ' The Logos has its birth through the interweaving of 
Forms with each other* 
References: Cornford, PTK 30??.; Ackrill, SPM 199^, Hackforth, CQ 
1945, 56-8; Bluck, JHS 1957, 181-6; Peck, CQ 1952, 32-56 and Phron. 
1962, 46-66; Lorenz and Mittelstrass, AGP 1966, 113-52; Hamlyn, PQ 
1955, 289-302; W. and M. Kneale, Development of Logic 20. 
This sentence has caused great difficulty, for at 263 a the statement 
'Theaetetus is sitting' is given as an example of a logos, yet it exhibits a 
combination not of Forms but of a single Form with a particular. There have 
been many attempts to solve this problem, and others are referred to in the 
discussions mentioned above. Cornford took P. to mean that ' at least one 
Form' must be used in every statement or judgement, but Ackrill was quick 
to point out that this is not what he says. Hackforth suggested that ???????? 
????? was different from ???????? ????? and that the ???? here are parts of 
speech, but neither Peck, Bluck nor Lorenz and Mittelstrass could believe 
this, the last-named pointing out that ???????? has occurred earlier, at 
240 d 1, where it is applied to the Forms Being and Not-being. Bluck, taking 
it as axiomatic that if all logos consists of a weaving together of Forms every 
statement must somehow involve at least two Forms, even if it is about an 
individual, concluded that the specimen logos about Theaetetus wove  
together the two Forms Man and Sitting. Hamlyn's solution is similar: 
Theaetetus * unpacks' into a list of all the Forms in which he partakes. So too 
Lorenz and Mittelstrass, but this is to treat rather light-heartedly the most 
puzzling feature of the passage: Theaetetus is not the Form of Man. 
'"Theaetetus is sitting" can be true', says Bluck (p. 182), 'because men are 
in fact capable of sitting.' But Plato says it is true, though it is not true either 
of all men or of the Form Man: to sit is neither an essential nor an exclusive 
attribute of mankind and has no place in its definition. The statement refers 
to Theaetetus alone. Peck was probably justified in calling Bluck's 'a 
desperate expedient', though it is interesting to note by the way how close it 
would bring Plato to Aristotle's epistemological theory of the perception of 
the universal (or specific form) through the individual. 'Though it is the 
individual that is perceived, perception is of the universal, e.g. of man, not 
Callias a man' {An. Post. iooai6-bi). 
Peck reversed Bluck's argument by laying it down that since the statement 
12 
161 
GHG 
Parmenidesy Theaetetus, Sophist, Politicus 
' Theaetetus is sitting' clearly does not make use of a combination of Forms, 
the sentence under discussion cannot mean that every logos is constructed by 
weaving Forms together. It only means that the combination of Forms is a 
necessary precondition of the existence of Logos because, to exist, it must 
combine with Being, and it is itself a Form B6035). The sentence is not 
intended as a definition, for at 260 a 7 the visitor says that having established, 
by the doctrine of the blending of Forms, that Logos can exist, it remains to 
agree on what it is. And when the definition comes, it is in terms of a  
combination of words, not of Forms. I have adopted Peck's interpretation as the least 
open to objection, though it may seem a little extravagant of Plato that in 
order to defend the possibility of Logos existing, he should cite the general 
conditions which secure the existence of anything at all. (' This ???????? 
includes the ???????? of ????? (or ?? ??) with all other ???? or ???? B59 a)', 
Peck, Phron. 1962, 57.) If unsatisfied, one can, I suppose, fall back on the 
simple admission of Martha Kneale that Plato * never dealt clearly with the 
distinction between singular and general statements'. 
(it) Republic 5 and the Sophist 
Stenzel represented a fairly widespread view with his statement (PMD 53) 
that when we find in Soph, that false opinion exists because '" Not-Being", 
to which it is related, exists', * This is in direct conflict with Republic, 478b, 
and Theaetetus, 189 a, b, where Not-Being was declared to be, not only 
unknowable, but inconceivable, because nothingness cannot be conceived.' 
(Cf. Grote, PL 11, 455 and Apelt, introd. p. 40.) But where is the conflict? 
In Rep. 5 we have at 477 a: * What fully is is fully knowable, what in no way 
is is entirely unknowable.' In Soph. B38c) Plato says (Cornford's  
translation) : * One cannot legitimately utter the words, or speak or think of that 
which just simply is not; it is unthinkable, not to be spoken of or uttered or 
expressed.' He removes the difficulty there by pointing out that the choice 
is not a simple one between * is' and * is not', because the verb * to be' is used 
in different senses. In Rep. 5 he is facing quite a different problem, the onto- 
logical status of the sensible world. Nevertheless he solves it similarly by 
positing an intermediate category which * shares in both being and non- 
being, and which it is not right to call purely and simply either' D79c). This 
is the object of doxa, lying between the known and the unknown D78 c). 
Plato has already escaped the Parmenidean dilemma. The Sophist's  
conception of ?? ?? ?? as ??????? is no doubt an advance, making possible the 
claim at 258d that 'we have not only demonstrated that "things that are not 
are", but have brought to light the real nature of "what is not"'. This is of 
162 
Politicus 
interest from the point of view of Plato's development, but the context is 
quite different and there is certainly no conflict. 
D) politicus1 
Introduction, outline and general remarks 
The position of the Politicus among the dialogues is plain: it is a  
continuation of the Sophist (p. 33). Some have attempted to date its 
composition by relating it to Plato's activities in Sicily. This is at best 
uncertain, though the date arrived at (between 367 and 362) is likely 
enough.2 The company is unchanged, but as respondent Theaetetus is 
replaced by the younger Socrates (to be referred to here as Y.S.).3 
For those tidy minds (mentioned in vol. iv, 130) which like every 
dialogue to have one single aim, one 'real subject', Haupt\weck and so 
on, Plato has for once provided an explicit clue. He describes the 
Politicus as primarily an essay in method: 'Has our search for the 
Statesman been proposed for its own sake, or rather to make us better 
reasoners on any subject?' ? Clearly the latter.' B85 d) Again at 286d: 
? Reason requires that we are content to give second place to an easy and 
quick solution of the problem we have set ourselves: our first and 
greatest care must be for the method itself, that is, learning to divide 
according to kinds.' A thorough examination of the method is indeed 
important, for as Plato will show, it can trap the unwary who apply it 
too mechanically, and demands an alert mind and constant use of good 
judgement if it is to be successful. On the other hand he does not say 
that the ostensible aim of tracking down the statesman is unimportant, 
only that in dealing with this or any other subject our natural desire for 
a quick and easy solution must not be satisfied at the expense of correct 
method, which would only mean that the answer when it came was 
wrong. One can understand therefore, what the dialogue shows plainly 
enough, that the primacy of method does not mean that the enquiry 
into statesmanship is a mere logical exercise, an illustrative example on a 
par with weaving here and angling in the Sophist. Rather it emphasizes 
1 The Pol. contains many interesting reff. to Athenian political, legal and economic practices, 
on which Skemp's introduction and notes may be recommended. 
2 See Skemp 14-17, with Tate's criticism in CR 1954, 115. (Reff. to Skemp are to his  
translation of the dialogue.) 3 On Y.S. see p. 63. 
163 
12-2 
ParmenideS) Theaetetus, Sophist^ Politicus 
the supreme importance of getting the definition right. Nor is the lesson 
of the Republic irrelevant, that the master of dialectic and the statesman 
are the same man. 
The Politicus has been called a 'weary' dialogue, but is not so for 
those who enjoy Plato's mastery of the art of weaving (the word  
imposes itself) different topics together, not offering us dry little treatises 
on logic, political theory or ethics, but passing from one to the other 
and back again in a natural process of thinking aloud, with the guiding 
mind of the discussion-leader not avoiding digression but ensuring that 
each topic has had its due before the end. 
An outline of the dialogue's structure will exhibit this. It sets out, like 
the Sophist, to define its subject through a diairesis, starting from a 
different division of knowledge. In the course of this, Y.S. is given a 
lesson in the dangers of a lopsided principle of division. It proceeds to 
its conclusion, which however is declared unsatisfactory because it has 
not distinguished the statesman from his nearest rivals. 
Next comes a long cosmic myth, introduced as ? relaxation' (paidia), 
but also to reveal the errors of treating the statesman as if he were a god, 
a being of a superior order to his charges, whereas he is only a man 
among men, and of failing to make a proper diairesis. 
A revised division follows, but is also rejected. To explain his objection 
the visitor will use an analogy, first however explaining and illustrating 
the use of analogy itself The chosen analogy is weaving in wool, which 
is now itself defined by a long diairesis, invaluable to those interested for 
its detailed information on Greek weaving technique.1 This is rejected 
on the same grounds as the analysis of statesmanship itself (thus  
showing up a defect in the latter), namely that it does not cut off weaving 
from some closely allied arts. A further, successful attempt introduces a 
distinction between principal and ancillary arts,2 one sort directly produc- 
1 Wilamowitz's remark is just (PI. i, 576 f., apropos of this and the passage on angling in 
Soph.): * It must not be forgotten that Plato had more observation at his disposal than he displays 
in his writings.' The other locus classicus on weaving for students of Greek technology is 
Aristoph. Lys. 567-87, the point of which is that weaving is an entirely feminine art; so it is all 
the more impressive that Plato should have the technical details at his finger-tips. 
2 ????? and ???????? 281 d, 287 b. In this connexion the distinction between causes and 
necessary conditions at Pho. 99a-b is often quoted. Closer in language at least are Phil, ??* and 
Tim. 46c, d. 
I64 
Politicus 
tive, the other providing instruments or means for the production. That 
all this is in fact analogy (rather than merely an example of correct 
division exercised on a simpler subject), because the weaver turns out 
to be cousin to the statesman as the angler was to the Sophist, will be 
explained a little later. 
The only part of the dialogue which could be called 'weary' is the 
long-drawn-out series of divisions leading to the definition of  
weaving,1 and for this the visitor immediately apologizes. But Plato has 
more than apologies in mind. In the guise of defending this wordiness 
he introduces the concept of the right mean. This in turn leads to yet 
another reminder of the importance of dialectic', the science of  
discerning both differences and affinities between groups of Forms or kinds 
(eide). 
Return to the statesman is effected B87 b) by an application of the new 
distinction between directly productive and instrumental arts, and an 
acknowledgement that dichotomous division is not always adequate: the 
division of instrumental arts must be sevenfold, and even then, the 
statesman's rivals are found elsewhere, among politically minded 
Sophists. The theme is now political theory. Statesmanship being a 
science, constitutions are classified according to the amount of  
knowledge they display. Ideally a state should be governed by a man or men 
of genuine knowledge, and nothing else would matter. Failing the ideal, 
a city's best safeguard lies in laws strictly enforced, but these remain a 
second-best because, being universally binding, they cannot do justice 
to the infinite variety of people and circumstances. As a 'side-issue' 
C02 b 8), the imperfect constitutions are ranked according to the 
tolerability of life under each. 
The final stage, to separate the statesman's art from others even 
closer to it than its counterfeits, is reached by yet another primary 
dichotomy of arts into (a) an art itself, e.g. rhetoric, the art of persuading, 
(b) a master-art of knowing whether an art should be learned and how 
and when applied. The statesman's knowledge gives him this priority 
over kindred arts also necessary to good government—generalship, the 
administration of law and justice, public speaking, education. These he 
will guide and direct towards his own end, which is to weave the whole 
1 It is conveniently tabulated by Ritter, Essence 239. 
165 
Parmenides, Theaetetus, Sophist, Politicus 
state, with all its institutions and every citizen—the bold and high- 
spirited with the quiet and gentle—into a close, firm and durable 
fabric. 
I have called this method of presentation enjoyable, but it is more. It 
points to connexions in Plato's mind which we might otherwise miss, 
especially those between method and subject-matter. The definition of 
weaving shows up a fault in the logic o( diairesis but is also a metaphor 
for the essential work of the statesman. The dichotomy of measurement 
into comparative and evaluative is an example of, and encouragement 
to, correct diairesis according to real kinds B85 a-b, 286 d), but the 
principle of division is axiological, a reminder that the Forms discovered 
by diairesis itself are not merely genera and species but patterns or norms. 
With so much established, we may look at certain topics in more 
detail, fortified against supposing that they were as segregated in Plato's 
mind as we now regard them for expository purposes. 
A) Logic and method 
(a) Collection and division. This is how the dialectical method of  
collection and division, already known from the Phaedrus and Sophist, is 
described once more in the Politicus. 
The proper procedure is, when perception first presents the common features 
of a number of things,1 to press on until one can see all their specific  
differences; and conversely when in a multitude of objects all sorts of disparities 
are detected, not to be shamed into giving up until one has penned all that 
are cognate into a single enclosure of similarity and included them within a 
genuine genus. B85 a-b) 
1 In the passages descriptive of this method here and at Soph. 253 d it is not easy to know 
whether Plato has in mind Forms or particulars as the starting-point. Perhaps he is not  
discriminating, as Hackforth suggested was the case at Phil. i6d. (See his PEP 23 n. 2.) Here 
however I have opted (unlike Skemp) for particulars, which are strongly suggested by the 
combination of ????? with ?????5. (Cf. Phil. \6??.) Even the dialectician must start (as in the 
Phaedo) from the evidence of his senses, making use of the universal human capacity to form 
elementary general concepts, and even at this early stage the non-philosopher may make mistakes, 
grouping, say, flowers by obvious but inessential differences which the trained botanist would 
ignore. I therefore give ???????? at b ? its ordinary reference to sense-perception, which is of 
course natural and easy, though not absolutely necessary. (At Soph. 253?? ???????????? is used 
with·' ??? as its object.) 
166 
Politicus 
This is a somewhat clearer exposition of the method than that at 
Soph. 253 d, quoted above on pp. 129 f. It is applied to' the way of  
statesmanship' at 258c: 'We must discover it, and separating it from other 
ways, stamp upon it its unique form.' In the Sophist the practice  
definition of the angler preceded that of the Sophist himself. Here Plato 
begins with the real subject, and only after failure turns for help to a 
parallel case. In the course of the long first diairesis the statesman appears 
as the nurturer of a hornless, wingless, tame two-footed herd incapable 
of interbreeding with other species. The summum genus chosen is again 
art (techne, dependent on knowledge), but a new initial dichotomy is 
adopted, into theoretical and practical. The statesman is assigned to the 
theoretical, not in the sense of being pure scientist or mathematician but 
like a master-builder who designs a house and supervises its  
construction, a worker with brain not hand. The risk of confusion is removed by 
dividing theoretical knowledge into critical (involving judgement alone 
with no action following) and directive. Knowledge is the sole  
qualification, and the man who has it deserves the title of statesman or king even 
if he remains a private citizen and only advises the actual ruler—the 
position Plato thought proper for himself and his associates.1 
The Politicus brings out even more clearly than the Sophist how far 
diairesis is from being a merely mechanical process. It is indeed 'easy 
enough to indicate but extremely hard to practice' (Phil. i6b-c). From 
the initial dichotomy onwards, every step involves personal judgement 
and choice, and in spite of apparent trivialities, sometimes introduced 
only to show how the inexpert may go astray, the continual watchful 
insight required to keep on the right track makes credible Plato's claim 
that dialectician and trained philosopher are one and the same. Both, 
one might say, are far advanced along the road to ' recollection' of the 
Forms. 'Acquisitive and productive' and 'theoretical and practical' are 
in his eyes equally legitimate divisions of the genus 'art', but not 
equally suitable for the investigation of a particular subject. It may be 
necessary to proceed quite far with the successive subdivisions before 
it becomes evident that something is wrong, that divisions are not 
1 259a. Cf. vol. iv, 23. He must also have remembered that in the Gorg. E21 d) he had 
described Socrates as the only one to practise the real art of politics. True, S. did not even advise 
the governing element (Apol. 31c), but in his eyes as in P.'s, where power lay with the demos 
there was no genuine ruler to accept advice. 
167 
Parmenides, Theaetetus, Sophist, Politicus 
being made 'at the natural joints' but haphazard, revealing only parts 
that do not correspond to genuine Forms B62 a-b). Then the  
successive steps must be retraced, to see at what point the error occurred 
B75 c-d). In the first diairesis, the rather laboured humour of hinting 
that man's closest competitor is the pig, and likening his mode of 
walking to 'the two-foot potency of the diagonal' B66b-c), as well as 
the ridiculous definition itself, are probably intended to emphasize the 
pitfalls of amateur dialectic. The failure of this first attempt is obviously 
contrived by the visitor with deliberate pedagogic intent. 
The most important rule is that the divisions of dialectic must  
correspond to reality, i.e. to the ontological distinctions between Form and 
Form. They are discovered, not imposed.1 The opposite fault comes to 
light when Y.S., in their search for the objects of the statesman's care, 
suggests dividing off men from beasts. This is to contrast a single 
species (Form) with an unanalysed aggregate. To put the whole non- 
human animal kingdom on one side of a dichotomy is not dialectical 
* division according to kinds' but the severance of a part which  
exemplifies no specific Form. Every Form is a portion of a wider genus, but 
not every portion is a Form.2 The visitor compares the Greek habit of 
dividing mankind into Greeks and foreigners (barbaroi, lit.  
inarticulate'), and the imaginary case of dividing number into 10,000 and the 
rest, inventing a collective name for numbers other than 10,000. On this 
principle cranes, who are reported to possess intelligence, might divide 
living creatures into cranes and beasts, including mankind in the latter.3 
1 ???????? ??? ???????, Phil. i6d. 
2 262b 1-3: 'Not to separate one small part over against many large ones, nor without regard 
to Form: the part must at the same time have a Form.' The second condition disposes of scholars 
(mentioned by Runciman, PLE 60 n. 1) who believe that P. rejects a division like that into 
Greek and non-Greek on purely extensional grounds. 
3 Many believe that besides the methodological point, P. here gets in a forceful criticism of the 
Greek attitude of superiority to the rest of mankind. Skemp A31 n. 1), though Tate (CR 1954, 
116) thought his view exaggerated, has the support of Friedlander (in, 287 f.), Ritter (N. Unters. 
77) and others. Field on the other hand denied it (P. and Contemps. 130 ?.). Skemp's view  
certainly goes dead against Rep. 470 b, as he admits, but he also points out that a different attitude 
was gaining ground among intellectuals of the fifth and fourth centuries. If there is a moral point 
intended, one would expect it to apply also to the example of the cranes, which would become 
a rebuke to mankind as a whole for vaunting its superiority over the beasts. A modern reader 
might naturally see it in that light, but for Plato man, with his immortal mind, his divine gift 
(Phil. 16 c) of forming general concepts, making it possible for him to work his way up to 
recognition of the Forms that he once saw, was indeed set apart from the beasts through his 
affinities with a higher world. 
168 
Politicus 
(b) The two types of measurement. ' More and less must be measured not 
only in relation to each other but also with a view to the achievement of 
a norm' B84b).1 
This is introduced as a principle of great importance, on a par with 
the distinction established in the Sophist between negative predication 
and denial of existence. It is no merely logical pronouncement, but 
reflects the whole Hellenic ethos of * nothing too much' and anticipates 
Aristotle's doctrine that goodness lies in a mean. It is 'what especially 
distinguishes bad men from good' B83c).2 The word metrion  
(translated 'norm' above), from metron, measure, is basically what is within 
measure, sometimes average' (Hdt. 2.32.6), but generally used 
approvingly of what is in due measure, moderate, sometimes almost a 
synonym for 'good'.3 
The first sort of measurement, Plato goes on, is sufficient for purely 
theoretical studies B84 a). The pure mathematician calculates the  
relation between numbers. (Ratio was basic to Pythagorean mathematics.) 
He works with arithmetical, geometrical and harmonic means, but there 
is no question of a right mean because he has no ulterior purpose in 
view. But both are demanded by pursuits which, like statesmanship, 
have a practical aim. Plato goes so far as to say that without the skill to 
judge excess and defect, politics, weaving and every other art would be 
destroyed B84 a), for all have to consider 'what is in due measure 
{metrion\ what is fitting, what opportune, what has to be done', and 
these are never found among extremes B84 c). 
All this has an oddly familiar air. In the Protagoras C37e-8b) Hippias 
recommends a moderate {metrion) length in speeches, as opposed to 
long harangues or excessive brevity. Prodicus laughed at orators who 
boasted of making very short or very long speeches at will: a good 
speech should be neither long nor short but metrion (Phdr. 267b). 
1 ???? ??? ??? ??????? ???????. There can be no philosophical significance in the change to 
?????? fiom ????? at 28363, though ????? there must mean real nature or essence. Ritter 
perhaps goes too far in claiming that the phrases ??? ???????? ???(? and ????? ?????????? at 
283 d 8 and e5 imply a deliberate modification of earlier doctrine about the opposition between 
being and becoming. (See his Essence, 183.) In its context the first phrase simply means that if 
anything is to be brought into being, then in the nature of the case the second kind of  
measurement is required. For a fuller comment in connexion with the Phil, see p. 233 n. 3 below. 
* ?????? are distinguished from ????? by their attainment of ? ??? ??????? ?????. At Arist. 
EN 1096325 ?????? in the category of quantity consists in ?? ???????. 
3 A typical example is Soph. O.C. 1212. 
169 
Parmenides, Theaetetus, Sophist, Politicus 
Conversely, when Socrates himself in the Protagoras, ostensibly  
advocating a hedonistic ethic, says that the good life can only be assured by a 
science of measurement able to estimate excess and defect,1 he confines 
his advice to the first kind of measurement, of pleasures and pains 
against each other, a process expressly condemned in the Phaedo (vol. 
iv, 234). Has not Plato himself asserted that the attributes of anything 
in this world (and it is with this world that the practical arts must come 
to terms) depend solely on its relation to other things? ' Large and small, 
light and heavy, can equally well be given the opposite epithet', what 
is double is also half and so on {Rep. 479 b). 
These statements need some sorting out. The Sophists were  
following the everyday use of metrion to mean not simply medial but of the 
right length. In a loose way it does measure things not against each 
other but against a standard, but they could not have defined their 
standard nor would they, with Plato at 286 d, have rejected pleasure as 
a legitimate consideration. The Sophists were relativists,2 and Ritter3 
does not fail to point in this connexion to Protagoras's teaching that the 
only metron is what appears to the individual (though according to the 
Theaetetus (p. 86 above) even he admitted that in practical judgements 
an individual man or state may be at fault). For Plato the standard is 
obviously provided by the changeless and definable Forms, culminating 
in the Form of the Good. After explaining the twofold division of 
measure, he immediately links it with diairesis, which, he repeats, is the 
method that enables one 'to divide according to Forms' B86d), and he 
1 See 356d-57b. ????????? . . . ????????* ??? ??????* at 357^ especially resembles the 
phraseology of the Pol. Cf. 283 en and e3, 285137-8. 
* For the two kinds, or degrees, of relativity in values see vol. in, 166, and for a modern 
treatment of the distinctions between objective and subjective, absolute and relative, Flew, Introd. 
81 ff. 
3 N. Unters. 87. All scholars except Ritter (including Skemp, p. 174 n. 1 and p. 79) have taken 
the ?????? at 28531 to be the Pythagoreans, but there is a strong case for R.'s belief that P. has 
in mind followers of Protagoras who upheld his doctrine of 'man the measure'. The case is 
strengthened (though R. does not mention it) by what we find at Rep. 454a about the ???????????, 
not dialecticians but eristics, who speak as they do ??? ?? ?? ???????? ???' ???? ???????????? 
?? ????????? ????????? v. (At Pol. 285 a the ?????? go wrong ??? ?? ?? ???' ???? ??????????? 
??????? ?????????vous.) And the ?????? here could well hold the theory of the relativity of 
sensation ascribed to ?????????? at Tht. 156a (pp. 77 ff. above). The Pythagoreans would then 
be those who rightly employ the first standard because as mathematicians they are concerned only 
with 'all arts which measure number, lengths, depths, breadths and velocities by relative standards' 
B84 e). From them the Sophists are distinguished as those who wrongly confine themselves to 
it in making practical and moral judgements. 
I7O 
Politicus 
warns that what has been said here will be needed when the time comes 
to demonstrate the very nature of truth.1 
But this is not the whole story. Though not Sophistic, the message 
of the Politicus is not an extreme absolutism.2 The discovery of the 
second type of measurement does not entail abandonment of the first 
B83610-11), and the enlargement of the due measure to include 
appropriateness, opportunity (kairos), what the situation demands,3 
vividly recalls Gorgias and his fellow-orators, for whom kairos in 
particular, the sense of occasion, was a prime requirement of successful 
speaking (vol. in, 272). In the middle dialogues the exaltation of the 
divine Forms, and man's status as an immortal soul acquainted with 
them in another existence, had thrust into the background an essential 
ingredient of Platonism and in particular of its legacy from Socrates, 
which leads here to a conception of the statesman and his task more 
realistic than the visionary creation of the Republic.4 In equating  
goodness with practical benefit Socrates saw that in ordinary life the good 
was always relative to a particular end, and as situations differed so did 
the good. Plato still shows him saying this in the Phaedrus: knowing 
how to administer different treatments does not make a man a doctor 
unless he also knows when and for which patients they are appropriate. 
In the Meno even virtues can be unwisely practised and lead to harm.5 
In the Phaedrus it is the rhetoricians who resemble the quack doctor, 
and their fault is the same as that of those in the Politicus who fail to use 
the second principle of measurement, namely ignorance of dialectic, 
which prevents them from even defining their own art and so knowing 
its true purpose {Phdr. 269b). Now the strands begin to combine. 
Central is the concept of purpose, function, an end (telos) to be 
achieved—it is to practical arts that the second principle applies—and 
the end exists objectively and determines the means. If you want to 
1 284c!. Skemp's expansion of the untranslatable ????$ ??? ???? ???? ?? ??????? ^?????? is 
* to give a full exposition of true accuracy in dialectical method \ Lit. ' to seek to demonstrate the 
accurate (or true) itself. Presumably the reference is to the Philosopher. 
2 Cf. Flew, Introd. 83: 'someone who believes that ethical values are objective can, with entire 
consistency, insist that the courses of conduct which these values determine must vary partly 
according to the particular occasions; and hence be relative to them'. Or Bambrough (RTG 143): 
'The objectivity of morality is compatible with its "situational relativity".' 
3 ?? ?????? ??? ??? ?????? ??? ?? ????, 284c 
4 This is expanded on pp. 183??. below. 
5 Phdr. 268a-c, Meno 88a-c. This side of S. has been treated fully in vol. m, 462-7. 
171 
Parmenides, Theaetetus, Sophist, Politicus 
produce a well-governed city or a woollen coat there are certain things 
you must do. It is ' in the necessary nature of production' B83 d). It was 
Socrates's insistence that the telos must first be understood, and as proof 
of understanding defined,1 that led to the doctrine of Forms. They were 
and remained ideals or standards to be aimed at (paradeigmata), even 
though human attempts must always fall short of their perfection. The 
further point that these standards lie in a mean between contraries is 
illustrated by the leitmotif of weaving with its firm warp and softer, more 
pliant weft, culminating in the final description of the statesman's own 
skill as that of reconciling bold and rash with quiet and peaceful  
temperaments; and that no earthly state can reach the ideal appears in the  
necessity for a permanent system of laws in default of the perfect ruler. 
Plato did not give up his belief in a universal Form of Good. Even 
Aristotle, who argued against it, could not believe that carpenters and 
shoemakers, and even parts of the body like hands and feet, should have 
their proper function and activities yet humanity as such have none, 
and he went on to describe this universal and exclusively human end. 
(See EN 1097 b 28-983 20.) Conversely Plato is coming closer to the 
Aristotelian position that for practical purposes knowledge of the 
highest Good is insufficient unless one knows what means to it  
(subdivisions of it?) are immediately applicable. The philosophic ruler, 
while not losing sight of the ultimate goal, must also be able to think 
things through to the proximate step 'which he himself can take'.2 His 
course resembles the 'downward path' of Republic 6, with the all- 
important difference that that 'involves nothing in the sensible world, 
but moves solely through Forms to Forms, and finishes with Forms' 
Eiib-c). The statesman must continue until he can grapple with 
situations in this space- and timebound world. To this development 
Plato's increasing use of the method of diairesis made an obvious con- 
1 The process in the practical arts is well expressed by Aristotle in the Ethics (? 111 \>i"?): 'The 
last step in the analysis is the first in production', and illustrated at Metaph. 1032b6-10: 'Health 
is produced by a train of thought like this: "Since this [of which I have the form (eidos) in my 
mind] is health, this must be present for the subject to be healthy, e.g. an equable state of the body; 
and if that, then warmth." So the doctor goes on thinking until he finally reaches a step which he 
himself can take... So in a way health is brought into being by health and a house by a house, the 
material from the non-material. For medicine and architecture are the form of health and house, 
and what I call substance without matter is the essence.' 
2 See previous note, and cf. Phil. 62a-b on the plight of the man who knows the Form of 
Justice but cannot use the tools of a trade or find his own way home. 
172 
Politicus 
tribution by bringing down the Forms as nearly as possible, within the 
necessary limits of knowledge, to the individual level. 
(c) The use of paradigm. To begin once again with Plato's own  
definition, a paradigm is used 'when one thing is rightly divined in another, 
separate thing and being brought into connexion with it brings about a 
single true opinion of each separately and both as a pair' B78 c).1 
His own use of the method makes this general statement clearer. 
Wool-working is obviously Other than' statesmanship, yet after 
recognizing within wool-working the contrasting arts of separation 
(carding) and combination (weaving) one can detect their reappearance 
in the organization of a community. It consists in a preliminary  
'carding' or combing-out of bad citizen-material from good through tests, 
followed by 'weaving' the rest together into a unified whole C08d-e). 
Thusparadeigma here, though often translated 'example', bears neither 
of the two usual senses of that word. It is neither an instance (as in 'an 
example of the classical style') nor a pattern to be followed ('he is an 
example to us all').2 Nearest to it is 'analogy'. In offering a 'paradigm 
of a paradigm' at 27yd9-278c2 Plato employs two senses. What he 
gives is an example of paradigm in the new sense in which he proposes 
to use it, taken from children learning their letters. They first learn to 
recognize them in short, easy syllables, then are shown others which 
they fail to read correctly. Then the teacher puts the familiar syllables 
besides the new ones, and points out where the same letters recur. This 
is precisely what Plato now calls the method of paradigm: the detection 
of common elements in different compounds. 
That he attaches great importance to it appears from 277 d: 
V. * It is difficult to demonstrate anything of importance without the use of 
paradigms. I suspect that in a dreamlike way we know everything, yet know 
nothing in reality.' 
Y.S. 'What do you mean?' 
V.' In an odd way I seem at this point to have raised the question of how 
we experience knowledge.' 
1 Reading ??? at 27806 with Friedlander {PL in, 527 n. 19). ?$ Skemp, following Campbell, 
Burnet and Dies. Both words have some MS authority. 
2 The meanings of paradeigma have been distinguished in vol. iv, 118 n. 2. 
173 
Parmenides, Theaetetus, Sophist, Politicus 
So paradigm has some relevance, albeit 'in an odd way', to the 
supreme question of philosophy, pursued in so many dialogues: What 
is knowledge and how do we know? Nor should this surprise us. 
Knowledge depends on the discovery of Forms, in which the first step 
is to recognize and isolate a common element, the 'single form' or One 
in and through all' of a number of different objects.1 The contrast 
between dreaming and waking was used in the Republic D76 c) to 
distinguish the mass of mankind, who are only aware of sensible things, 
from the philosopher trained to see beyond them to the Forms whose 
being they share; and again in the Meno (85 c) to describe the state of 
mind of one whose progress from belief to knowledge is still  
incomplete. There and in the Phaedo the progress was towards recollection of 
pre-natal knowledge, and the fact that Plato can remind us so  
unmistakably of these dialogues without mention of this once central doctrine 
is symptomatic of the shift of interest in the present series towards a less 
metaphysical treatment of epistemological questions.2 
The method as Plato introduces it here could be disconcerting, for if 
it were intended to serve the same purpose as definition in the Meno we 
should have to suppose that combination and separation represented 
the actual eidos (form) of weaving and government, which would  
therefore belong to the same eidos (species). This would hardly conform to a 
dialectical collection followed by division 'according to Forms', nor 
would paradeigma be a suitable term for that. Plato warns us that its 
relation to the acquisition of knowledge is 'curious' or 'unexpected' 
(???* ?????? 277 d 6). But something essential to a certain art can, mutatis 
mutandis, be revealed by an analogy taken from a quite different genre; 
and 'analogy' is the English word which comes nearest to paradeigma 
1 For these and similar expressions in Meno see vol. in, 433 n. 1. 
2 Ritter's explanation of P.'s silence about ???????? was that he had always intended it 
metaphorically and found that he had been grossly misunderstood. (See his N. L/nters. 80-2, and 
for the metaphorical interpretation Essence 121-3.) Others have argued that he once believed in 
it but dialectic has now replaced it. Yet as Rees pointed out in Proc. Ar. Soc., Suppl. vol. 37, 
i72fT., the two combine happily in Phaedrus, and we should not assume that P. has abandoned 
the doctrine altogether, any more than he has abandoned the conception of the Forms as 'bodiless, 
fairest and greatest' {Pol. 286a). See Gulley's cogent arguments in CQ 1954, 209-13. 
In the section on collection and division I have already spoken of'Forms' to emphasize P.'s 
point that the method must have an ontological basis; but he says nothing there inconsistent with 
supposing that the objective realities which he insists on might be within the nature of the 
phenomena themselves. 
174 
Politicus 
in this context. Perhaps one might say that the paradigmatic method 
itself provides a paradigm of (i.e. is analogous to) the philosopher's 
pursuit of knowledge, and is itself a valuable epistemological tool. 
B) Forms in the ' Politicus' 
The words eidos and idea existed before Plato appropriated them for his 
transcendent, intelligible patterns of sensible things and actions, and he 
himself continued to use them frequently to mean no more than kinds 
or species, and sometimes even in their root meanings of outward 
appearance. Hence the question whether, in using these or analogous 
terms in the later dialogues, he has at the back of his mind the exalted, 
other-worldly beings of the earlier, is and will probably remain a 
matter of controversy. In passages exemplifying the method of 
diairesis, the status of the eide in the phrase 'to divide according to eide9 
is not easy to decide. To treat them as nothing more than kinds or 
varieties certainly suits the context. Has Plato then ceased to believe in 
those divine entities on which he lavished such eulogies in Phaedo and 
Phaedrus? Has he, as it were, lost his faith in the 'place beyond the 
heavens', or is it simply that this aspect of the eide has less relevance to 
his present concern for correct philosophical method?1 This is the nub 
of the controversy between the so-called unitarians and their opponents, 
and such is the spell of Plato that an element of apriorism inevitably, if 
sometimes scarcely consciously, enters the argument: Plato was a great 
philosopher, and this he could not have been if all his life he believed 
(or alternatively did not believe) this or that, according to the  
interpreter's notion of what in a philosopher is sensible and desirable. My 
own position should be clear by now. Plato was one of the greatest 
philosophers, and that largely because he combined, simultaneously and 
uniquely, dialectical skill with a metaphysical, indeed religious belief in 
a supra-sensible realm of divine essences, and came nearer than anyone 
else to relating it rationally to the world of human experience. 
Two passages have been thought specially relevant to the discussion. 
(i) 278c-d. This immediately follows Plato's explanation of the 
use of paradigm by the example of teaching children to read, and he 
1 I have suggested that in the Soph, it was even an impediment. This section should be read in 
conjunction with pp. 161 f. 
175 
Parmenides, Theaetetus, Sophist, Politicus 
continues the metaphor1 whereby letters and syllables stand for 
elements and compounds. With some hesitation I offer my own 
rendering of this difficult passage. 
Should we be surprised then if our mind naturally undergoes the same 
experience [sc. as children learning to read by the paradigmatic method] in 
connexion with the elements [letters] of all things, and at one time, and in 
certain cases, guided by the truth, stands firm about each single one of them, 
but again, in different cases, is confused about them all? Some of the  
constituents it somehow guesses correctly in the compounds themselves, yet it 
fails to recognize the very same elements when they are transferred to the 
long and difficult syllables of everyday life.2 
Skemp and Campbell (ad loc.) both take 'the elements of all things' to 
be the Forms.3 They are certainly abstractions like combination and 
separation, but an opponent could justly claim to find no evidence here 
of the Forms of the middle period. Plato is now bringing the method of 
paradigm, just as he has illustrated and described it, to bear on his 
present subject. That subject is the practical arts (technai), and he is 
making the point that there are certain basic and elemental skills which 
enter into widely different occupations. These are the elements referred 
to here. In some cases (weaving) they are obvious, but in others, amid 
the distractions of ordinary practical life with which a complex art like 
politics is concerned, analogous skills are overlaid and concealed. If the 
would-be statesman can grasp the connexion—see, that is, through the 
simpler analogy of making a cloth of warp and woof, that his primary 
task is to reconcile contrasting human temperaments in a stable  
community based on consent (cf. especially 3ioe)—he will tackle the  
confusion of day-to-day political life with an eye fixed on the essential, 
ultimate aim. Such is the use of paradigm. It calls for no overt reference 
1 If it was a metaphor. Aristotle (Metaph. 998323) speaks of ????? ????????, i.e. the simple 
sounds composing vocal utterances (of which letters are the symbols); but P. was said by 
Eudemus (ap. Simpl. Phys. 7.14 Diels) to have been the first to apply the word to the elements of 
physical and generated things. Skemp ad loc. (p. 161 n. 1) produces no solid evidence that it had 
been so used earlier. 
* ??? ?????????. This can mean concrete things, or alternatively circumstances, affairs 
(especially political affairs), and trouble or annoyance B8563). 
3 C. compares Tht. 201 ff., where however ???????? are the elements of physical things as 
Cornford said (PTK 143 n. 1). 
I76 
Politicus 
to transcendent pattern-Forms, and Plato makes none.1 Here he is 
primarily concerned with the application of a certain type of intellectual 
insight to the achievement of practical goals, and his discussion of 
statesmanship is on a more mundane level than that in the Republic. 
The absence from it of the Forms need not necessarily imply their 
abandonment if evidence can be found to the contrary. Indeed I have 
ventured to suggest that in extending the use of paradigm  
(unnecessarily for his present purpose) to the whole problem of knowledge 
Plato has in mind the existence of Forms as its basis. But if not denied, 
neither is this confirmed by the present passage. 
(ii) 285d-6b.2 This follows Plato's assertion that the object of their 
enquiry is not merely to define the statesman but equally to assist their 
mastery of dialectic in general. 
And as for weaving, no one in his senses would want to pursue its definition 
for its own sake; but there is something which I believe has escaped most 
people's notice. Some real things have perceptible likenesses which can easily 
be grasped.3 There is no difficulty in pointing them out when one is asked for 
a definition and wants an easy and trouble-free method of exhibiting them 
without words. But the greatest and most precious things have no image so 
wrought as to be manifest to men, which the man who wants to give his 
questioner full satisfaction can point out to him, and by impressing it on one 
of his senses adequately satisfy his mind. Therefore we must train ourselves 
to be capable of giving and accepting a verbal account of everything; for the 
things that are bodiless, being the finest and greatest, can be clearly shown in 
words4 alone—nothing else—and everything that we are saying now is said 
for their sake. But practice is always more easily exercised on the lesser than 
on the greater. 
1 Skemp assumes (p. 162) that P.'s thought here is metaphysical as well as logical:  
Statesmanship and Weaving, as Forms, are the complex 'syllables' whose 'letters' are 'the more 
general and universal Forms in which they partake', e.g. Combination and Separation. I  
sympathize with this view, but on the strength of the passage here quoted it cannot be said to be 
proved. 
* For an alternative version see Owen in Ex. and Arg., 35of. 
3 Reading jbaSicos with the MSS. See Skemp's and Owen's notes (the latter in Ex. and Arg. 
350 n. 3). Skemp's choice of ??????? is surprising. He believes ?? ???? to be the Forms, but 
it is hardly Platonic to call these |!>???? ?????????? . To grasp the Forms is reserved for the 
few, after arduous philosophical preparation. The balance of the sentence is also against the 
emendation. 
4 The Greek word is logos, with all its manifold meanings: statement, definition, argument and 
much else. Perhaps Campbell's 'rational account' or Owen's 'explanation' would be better, but 
here P. seems to have had especially in mind the distinction between visual representation and 
verbal description. 
13 
177 
GHG 
Parmenides, Theaetetus, Sophist, Politicus 
It has been generally assumed (and certainly by myself) that this 
passage contained a clear reference to Plato's * middle-period theory of 
paradigm-forms'. Recently however G. E. L. Owen has denied this in 
a closely-reasoned article of which all unitarians must take note.1 Plato, 
he argues, is still defending the tedious length of his analysis of weaving. 
At 277 c he has emphasized the superiority of verbal explanation to 
visual aids whenever a pupil is able to follow them. Here he simply adds 
that the most important subjects cannot be shown in visible images at 
all. This is meant literally. The images are still man-made: weaving can 
be exhibited pictorially but statesmanship cannot. The words 'bodiless, 
finest and greatest' at once suggest Forms to one acquainted with the 
Phaedo or Symposium, but Owen notes that at Rep. 599c 'greatest and 
finest' is applied to 'wars, military commands, the government of cities 
and the education of men'.2 
Others have seen this passage as an obvious parallel to Phdr. 25ob-d, 
where Plato says that whereas earthly copies of the Form Beauty are 
visible, Justice, Sophrosyne 'and other things precious to the soul'  
produce no such immediate sensuous impressions, and only a few can 
apprehend the originals through such images as we have. Owen's reply 
is to dismiss the Phaedrus passage as mere myth and poetry, not 
seriously meant (p. 349). He points out that Socrates himself, in his 
concluding prayer to Eros B57a), describes the language of his whole 
palinode as poetical and says that, except for its introduction of the 
method of collection and division, it should be regarded as paidia. 
Plato's use of this word has been discussed in vol. iv,3 and here it is 
sufficient to note that at 278 b he applies it to their whole conversation, 
1 'P. on the Undepictable' in Exegesis and Argument (Studies . . . Vlastos, 1973). I must omit 
many interesting points not central to the present argument, but pp. 354^ on the meaning of 
????????? should not be passed over. In Crat. the object described as ???????? ?? ? ???????? 
?????3??? is not a tool made by man. If it were, the operator would model his new shuttle on the 
broken one, which he is expressly said not to do. He models it on ?? ?????, which deserves the 
name of 'really existing shuttle* (???? ? ????? ??????) more than any shuttle ever made. I 
cannot agree that O.'s point 'is unaffected by this* C55 n. 14). 
2 Not, however, ???????, and O. passes rather lightly over this word on p. 356. Cf. not only 
Symp. 211 a 5-7, 211 e-i2a (where moreover ?????? is used of the earthly imitations of Beauty), 
but also the Politicus itself, at 269 d. If however Rep. may be cited in evidence that middle-period 
Forms are absent from this particular passage, it is no evidence that they have been abandoned, 
for the existence of transcendent Forms is a central feature of that dialogue. 
3 See especially pp. 60-3, and for my own interpretation of the Phdr. as a whole, ch. viC) of 
that volume. 
178 
Politicus 
including the practical psychology of rhetoric' with which Owen  
contrasts the more lyrical portions. It is in the same dialogue that he extends 
it to every written composition B76 d, 277c). Even at 265 b his Socrates 
says (with his usual eironeia) that the myth * may have attained a measure 
of truth' and that 'blended with it was a logos that had some claim to 
credibility'. He adds B65 d) that the myth itself has exemplified the 
dialectical method, thus giving 'clarity and consistency' to the  
definition of love which it contained. To understand Plato one must  
recognize that he may present the same doctrine in myth and again in dialectical 
argument. If one denies that the poetical and religious language of the 
myth conveys philosophical truth for Plato, where is one to stop? Much 
of the Symposium, and what is given as argument in the Phaedo, will 
have to be discarded. What of immortality itself, the basis of all the 
rest,1 which has been seriously propounded outside the framework 
of a myth in both Phaedrus (the argument from self-motion) and 
Phaedo? 
Owen lays stress on the context of our passage, but the immediate 
context is the admonition that the aim of the enquiry is not even to 
define statesmanship (let alone weaving) but to become better  
philosophers. This is surely a warning that what follows will for a moment 
lift the argument from its immediate subject to a reminder of first 
principles, which if Plato still believed in them would be the Forms. All 
this does not prove that the Forms are referred to here, but if I am 
wrong and they are not (as I have argued may be the case at 278 c-d),2 
this in turn is no proof that Plato no longer acknowledged them. They 
are plainly referred to, in their 'middle-period' form, at 269d: 
To be always unchanging and constant3 belongs solely to the most divine of 
all things, and body is not in that class, What we call the universe or cosmos 
has received many blessed gifts from its creator, but nevertheless it partakes 
of body and cannot therefore be utterly exempt from change, though its 
motion is as far as possible constant, uniform and in the same place. 
If this occurred in a 'middle' dialogue it would be taken for granted 
1 For the intimate connexion between the immortality of the soul and the existence of Forms, 
see Pho. j6d-e, 
* May one occasionally leave a decision to the reader? It is certainly to be hoped that all 
readers of this volume will be readers of Plato. 
3 ?? ???? ????? ??? ??????? ?????, P.'s regular description of the Forms. See p. 141 n. 2. 
179 
13-2 
Parmenides, Theaetetus, Sophist, Politicus 
that it described the divine (Pko. 8ob etc.) and unchanging realm of 
Forms, and its relation to the physical world, so why not now? 
Another pretty clear reference to Forms as paradigms is at 300c. In 
an ideal world a philosopher-statesman would act on his own initiative, 
in the light of his knowledge, without written laws. Next best however 
is a written code to which everyone must conform. * These written laws 
would be in each case copies of reality,1 if they are composed on the 
instructions of those who know.' The philosopher, that is, will model 
the laws on the changeless, perfect moral Forms which he alone  
remembers. They will be direct copies, not at two removes like the copies 
condemned in the discussion of mimetic poetry in Rep. 10. Plato still 
thinks of the philosopher-statesman as he did in Rep. 6 E00 c, d) as one 
who * beholds things that are unchanging . . . completely orderly and 
rational'. * These he imitates' and Studies to implant them in human 
behaviour both private and public.' With this must be taken the  
references to the one true, right or godlike constitution, the paradigmatic 
Form of a state of which all human states are imitations, some better, 
some worse B97c). Plato is explicit that the perfect statesman, and 
therefore the ideal state, do not exist on earth. 
Since there is nowhere to be found, as we assert, a Royal Being in our cities 
like the Queen in a hive, one man outstanding both in body and mind, we 
must, it seems, meet together and draw up laws, pursuing the tracks of the 
true constitution. C01 d-e) 
And at 303 b: 'For that city is to be distinguished from all others, as 
god from man.' Finally, Sophist and Politicus are undoubtedly products 
of the same stage of Plato's development, and it would be strange indeed 
if a renunciation of the Forms came between them. Yet I hope it has 
been shown in the previous chapter that the Forms are still present in 
the Sophist. 
C) The myth2 
Something has gone wrong with the definition of a statesman, and as 
a respite from strenuous dialectical exercise the visitor undertakes to 
1 ???????? ??? ????????. The best expansion of this brief expression is in Aristotle's Pro- 
trepticus. See vol. iv, 548. ol ??????? would be people like P. himself and the members of the 
Academy who did in fact act as legislators for a number of states (vol. iv, 23). 
2 I have collected in an appendix (pp. 193-6) some evidence for its sources, which may 
interest the historically-minded. Skemp has a long excursus on the myth (pp. 81-111) with many 
refT. to modern writings. 
l8o 
Politicus 
uncover the mistake through a story.1 Many ancient myths, he begins, 
contain a dim folk-memory of the same historical event. He instances 
the reversal of the course of sun and stars by Zeus in the myth of 
Atreus and Thyestes,2 the age of Cronus, and the earthborn men. All 
these reflect the fact that the universe suffers a periodical reversal of its 
rotation. In one era God controls the motion, but he cannot do so for 
ever because its material embodiment prevents it from either being 
motionless or enjoying a perpetual, single and perfect motion. He does 
his best for it, being perfectly good himself, nor must we suppose that 
there are two opposed gods turning it in opposite directions. What 
happens is that when God lets go, it begins to turn itself in the opposite 
direction, for (as in the Timaeus) it is a living and thinking creature.3 
The moment of reversal brings terrible convulsions, destroying much 
of the human race. 
Then (i) in the period opposed to ours, when the sun travels from 
west to east, the sequence of life is first halted, then reversed. Men and 
animals grow younger, white hair darkens, adults dwindle to infants 
and finally disappear into the earth, whence the next generation are 
born fully adult.4 This is the fabled age of Cronus, with all its  
traditional features: no fierce beasts, no wars or factions, earth yielding 
food untilled, perfect climate making clothes and shelter unnecessary. 
Moreover, to each tribe of animals, including mankind, is assigned a 
1 For its serious purpose see 274c· 'Here we may end our story, and turn it to account in 
discovering what a mistake we made when we demonstrated the royal and statesmanlike character 
in the previous argument.' 
2 Known from Euripides, Or. 1001-6, El. 726-44,1.T. 192-5. 
3 209d7-c2. The motion is not mechanically caused, as 'momentum* in Skemp's translation 
of 270a5-7 might suggest, though physical characteristics—size, equilibrium and a tiny pivot— 
provide the necessary conditions like the bones and sinews of Pho. 98c~99b, or the ???????? of 
the present dialogue. 
4 A nice point: are they born old or in the prime of life? The evidence is conflicting. If the 
course of life is reversed, they should logically start in old age; and at 271b4 P. says tous 
????????? ??? ??? ??? ?????? ?? ??? ????? (though this might refer to the generation already 
old at the transition). So Frutiger, Mythes 242, and Koster, Mythe de P., de Zurathoustra et des 
Chaldeens 45: 'on nait vieillard*. On the other hand Skemp has (p. 153) 'stalwart in their prime 
of life* (which however is not in the Greek he is translating), 'in the prime of adult life* (no) 
and 'in P.'s era of Kronos there were no old men* A11). This would correspond to the earth-born 
warriors in the traditional myths of Cadmus and Jason, and to the Giants. Moreover Adam {Rep. 
vol. 11, 297), who favoured the other view, supports it by a mistranslation of ????? ????? at 273 b 
(which however he had strongly defended in CR 1891, 445), where P. is in fact describing the 
beginning of the other era (our own). Yet Hesiod spoke of a time (surely not unconnected with 
the source of P.'s idea) 'when men are born grey-headed* (Erga 181). 
181 
Parmenides, Theaetetus, Sophist, Politicus 
minor god,1 under whose direct supervision no political organization 
is necessary. Nor was there marriage or begetting when all were born 
from the earth. At this point the visitor pauses to enquire whether the 
men of this (as it was usually thought) golden age were in truth happier 
than ourselves. The answer is Platonic. Yes, if they used their leisure 
and other advantages (including the gift of conversing with the animals) 
to acquire wisdom, but not if, as the legends suggest, they filled  
themselves with food and drink and wasted their time in idle talk.2 
B) With the world's latest reversal, ushering in the present era, * the 
purpose of the story' comes into view B72d 5). God relinquished his 
control, and the subordinate deities followed suit. After the inevitable 
upheavals and loss of life the universe settled down to guide itself on the 
lines which it had learned from its maker, but the imperfections  
inherent in corporeality have caused it to forget these more and more, and 
disorder and chaos are growing as the era approaches its end. Then God 
will again take charge and save it from complete destruction. However, 
our present concern is with mankind and the ideal ruler B7365). After 
the cosmic reversal, life progressed from infancy to old age and  
conception by sexual means replaced birth from the ground. * Now comes the 
point of the whole tale' B74b!). Bereft of divine care, and with nature 
turned hostile, men became a prey to wild beasts, and were without arts 
or tools to provide for themselves the necessities which the earth once 
yielded spontaneously. Their miseries^ were lightened only by the 
divine gifts and instruction of which tradition tells: fire from  
Prometheus, technical skill from Hephaestus and Athena, agriculture from 
other gods. Above all, from then to now mankind has been left to 
manage its own affairs and look after itself. 
The declared lessons of the myth (apart from relaxation) are (i) that 
our statesmen are only human: the days are past when we were ruled 
by gods; (ii) that the assimilation of statesman to herdsman had led to 
1 Adam's' God himself was the shepherd of the earth-born* (o.c. 296) must rest on a  
mistranslation of 271 e5-6, which means 'A god supervised them personally.' The supreme god has  
oversight of the whole cosmic motion, and allots to each subordinate his province, as in Timaeus 
DibfT.) he leaves to them the creation of mortal beings. 
2 A number of jokes in Aristophanes show that 'Cronian* could mean old-fashioned to the 
point of stupidity. For reff. see Baldry, CQ 1952, 85. 
3 Described in greater detail in Protagoras's myth of human progress, where too emphasis is 
laid on the lack of the political art (Prot. 32^4-5, 322b 5). 
182 
Politicus 
a faulty diairesis, which by implying responsibility for the actual  
nutriment of his flock, failed to divide him off from such classes as farmers, 
traders, millers and bakers, doctors and trainers. To many the length 
and elaboration of the story have seemed disproportionate to this 
simple task, and they have sought something more recondite. Best is the 
conclusion of Solmsen,1 which not only takes into account the doubts 
cast on the wisdom and happiness of life under Cronus, but is in  
keeping with the ethos of this whole group of dialogues. The myth shows 
that change and even deterioration are essential phases of a corporeal 
world, and that 'the philosopher who confines his attention to the 
Permanent and Unchanging misses a great deal, and cannot arrive at an 
integrated picture of the world'. In saying this we need not quarrel with 
the dictum of Wilamowitz (PL I, 576) that though in the circumstances 
to look for hidden meanings is pardonable, it is more cautious and 
truer to content ourselves with the welcome fact that Plato still takes 
pleasure in telling stories. 
D) Political theory 
(a) Politicus and Republic. Reading the Politicus cannot fail to bring to 
mind the earlier Politeia and the question how far Plato's standpoint has 
changed between the two dialogues. Both are based on the fundamental 
Socratic principle that government is an art, a techne dependent on 
knowledge,2 which few if any can master B97 b-c). The familiar Socratic 
analogies with other technai are prominent, for instance in the  
statesman's3 attitude to rules or laws: the skilled captain needs no rules but 
his own techne, his techne \% superior to rules B97a); the doctor may do 
better for his patient by relying on his own techne than by sticking to 
what the books prescribe B96 b). Yet with this insistence on an art to be 
learned, the Politicus is silent on the all-important subject of education 
for statesmanship which occupies the central place in the Republic. 
Evidently Plato had no wish to change the programme there laid down, 
1 P.'s Th. 85 f. I admit to having slightly altered it, preferring not to speak of change and 
deterioration 'in Reality*. 
2 Socratic: see vol. in, 409ff. For the repeated use of ????? in this connexion, e.g. 297a and b, 
300CI0. 
3 Or 'king*. In this dialogue 'statesman* and 'king* are convertible terms for the ideal ruler 
B76 c 8, e 13). The king, of course, is at the opposite pole from the tyrant, who governs for purely 
selfish ends. 
183 
Parmenides, Theaetetus, Sophist, Politicus 
and saw no point in repeating a task already done. Knowledge of  
goodness, to be imparted to others as true belief C09 c), is a prime requisite 
in both. 
The Politicus reaffirms the Republic's distinction between a single 
ideal polity and all others, but whereas the Republic concentrates on the 
ideal, whether or not Plato thought it could ever be realized, the 
Politicus recognizes that it is not of this world. The best of our  
statesmen are only human, and the present aim is, without losing sight of the 
'one true constitution' as standard and guide, to plan a society as like it 
as human imperfections allow. Plato sometimes seems to have three 
levels in mind: (i) the ideal statesman, god rather than man, whose  
enlightened will is his only law; (ii) the best sort of human statesman or 
political reformer (who himself has not yet appeared), whose  
qualifications and policies form the main subject of enquiry in the Politicus; 
(iii) the political Sophist or spurious statesman, who pretends to the art 
and passes himself off as a statesman instead of the mere party-hack that 
he is C03 b-c). This includes all contemporary politicians.1 But the 
distinction between (i) and (ii) is not always clear, and he admits at 
301 a-b that he uses the same title for both. 'When one man rules 
according to laws, imitating the one with knowledge, we call him king, 
making no distinction of name between monarchy with knowledge and 
constitutional monarchy with right belief.' This leads him into at least 
apparent contradiction on the use of a written code of laws and the 
desirability of government by consent of the governed. More points for 
comparison with the Republic will arise as we proceed. 
(b) Rule by force or consent. After the myth Plato points out certain 
grave defects in their previous diairesis. One is that by omitting the 
distinction between enforced and willing submission they confused two 
very dissimilar types, the king and the tyrant. Oversight freely accepted 
belongs to 'the genuine king and statesman' B76d-e). According to 
Xenophon, this distinction went back to Socrates:2 'In his view the rule 
of men with their consent and according to the laws of the state was 
1 The Critias, which has several elements in common with the myth of the Pol., casts a similar 
threefold classification into mythical form. See Campbell, Pol. xlviii. 
2 Xen. Mem. 4.6.12 (vol. m, 412). According to this passage Socrates foreshadowed the 
division of constitutions at Pol. 300eflf. (p. 188 below). 
184 
Politicus 
kingship, but rule over unwilling subjects, not according to law but at 
the whim of the ruler, was tyranny.' Somehow Plato must have  
reconciled this in his mind with what he says after enumerating the recognized 
types of constitution—constitutional monarchy and tyranny,  
aristocracy and oligarchy, democracy—distinguished by the number who 
govern, their wealth, and the consent or otherwise of the governed. 
Since government is a branch of science, he goes on B92 b-c), the 
criterion should be none of these, but solely the mastery or otherwise of 
that science. Given the requisite knowledge, which few if any can attain, 
it makes no difference whether the subjects bow willingly to his rule, or 
whether it is with or without a code of laws. He may at his discretion 
purge the city by banishment or executions, or increase it by  
immigration. Where cautery or surgery is needed, the doctor will best serve his 
patient's interests by carrying on regardless of his cries and protests. 
Even before starting his reforms the statesman will, like a good 
craftsman, reject any bad material. That is, after a series of tests he will 
put to death, expel or degrade any who prove incapable of acquiring the 
social virtues Co8c~309a). This corresponds to the 'cleaning of the 
canvas' in the Rep. E01 a) and, interestingly enough, to what  
Protagoras laid down in the Prot. C22 d) as a prerequisite of life in zpolis. Any 
laws will have been drawn up by the scientific statesman himself and 
may be altered as he thinks fit. The current assumption that he must 
first gain the agreement of the citizens, though admittedly plausible, is 
wrong. It seems a far cry from the Crito^ where Socrates, whom Plato 
regarded as the one scientific statesman, laid down as the only allowable 
alternatives obedience to the laws or their amendment by peaceful 
persuasion.1 
It is, however, the ideal statesman who is here portrayed. The only 
modification required in the earlier statement is that the essential  
difference between him and the tyrant is one of motive. The statesman 
pursues justice, truth and the welfare of society, the tyrant his own and 
his friends' aggrandisement and the destruction of his personal enemies. 
Since this is what men have learned to associate with absolute power, 
they will entrust it to no one. Nevertheless if the one true statesman 
should ever appear they would recognize and welcome him and leave 
1 Crito 5ie~52a. For S. as sole practitioner of the art of statesmanship see Gorg. 521c!. 
185 
Parmenides, Theaetetus, Sophist, Politicus 
him to inaugurate unopposed the one true and happy form of society.1 
The question of force or consent would not arise. 
(c) The role of law in government. When the visitor maintains that the 
perfect state, whose ruler acts with knowledge and justice, will need no 
laws, and that law-abiding states, though better than some, are only 
imitations, even the docile Y.S. rebels (' The rest of what you say sounds 
reasonable, but that one should rule without laws is a hard saying'),2 
and the visitor agrees to discuss further the Tightness of ruling without 
laws. He is emphatic that the scientific ruler will often be guided by his 
knowledge without paying attention to any written code. Any code 
must be universal and indiscriminate, a blunt instrument (????????? 
295 a) which can never provide for the individuality and variety of 
human personalities, needs and circumstances. 'It cannot be right for 
what is everywhere uniform to deal with what is never uniform' B94c). 
This is followed by a violently sarcastic attack on laws and their  
universal enforcement. Admittedly a ruler who was a law to himself could do 
great harm. So too a doctor might poison his patient for money, or 
a captain maroon his passengers or throw them overboard, and Plato 
amuses himself with a satirical picture of what it would be like if for 
that reason medicine and navigation were entrusted to untrained men 
following a code of legally binding instructions. The Assembly would 
welcome the advice of laymen on the use of medical drugs and  
appliances and on seamanship, and officers would be appointed annually 
by lot to perform cures and command ships strictly by the book. 
Further provisions include a ban on research into medicine, navigation 
and their ancillary sciences. Anyone undertaking it will be denounced 
as a chattering, star-gazing Sophist, and may be brought to trial as 
a corrupter of the young. In this caricature of Athenian democracy, 
with its obvious references to Socrates,3 he carries to its logical  
conclusion the soberly critical description of it in the Protagoras Ci9b-d). 
1 A similar hope was expressed in Rep. about the public's reception of the philosopher-king, 
a phenomenon as yet beyond their experience D98d~502a). 
* 293 a. The Greeks took pride in owning no master but the law. (See vol. in, 69f.) Y.S.'s only 
other doubt concerns the idea that the laws might be altered without popular consent B96 a), 
another affront to Athenian democratic theory. 
3 For Socrates as a * star-gazing Sophist* see vol. in, 364 and 374; iv, 431 n. 3 and 499 n. 4. 
186 
Politicus 
Since the Assembly listens to the pronouncements of laymen on the 
supreme and all-important science of statemanship, why should it not 
pay equal heed to their views on other arts and sciences? 
After this outburst Plato unexpectedly turns round and declares that 
after all we cannot have the ideal. No ruler can be everywhere at once 
and attend to every case. (He always speaks in the singular. The  
possibility of delegating legislation to subordinate ministers or committees is 
ruled out by the impossibility of finding enough men qualified in the 
political techne.) Therefore written laws with all their defects must be 
adopted as a second-best. Here is another departure from the Republic 
in the direction of practical politics. There he had dismissed detailed 
legislation as useless in either a well or a badly constructed state D27 a). 
Here he argues that although a burden of petty legislation might destroy 
all arts and make life intolerable, there is something worse still, namely 
that the official charged (perhaps by lot) with administering it should 
prove corrupt as well as ignorant, and abuse his position to gain  
personal power. Plato's lasting idealism is at odds with his present  
determination to be practical. He has just argued that fear of tyrants should 
not debar the expert statesman from dispensing with laws, any more 
than the occasional appearance of an unscrupulous doctor justifies  
replacing the expert's knowledge by a state code of therapeutical practice. 
Now he says that disobedience to the laws, founded as they are on long 
experience and public approval,1 would inflict much more harm on 
society than the existence of the laws themselves. The Politicus  
forecasts the detailed legal enactments of the Laws, where they are still 
described as a second-best for the same reasons as are given here. (See 
Laws 875c-d, translated on p. 335.) 
The true statesman, then, will be right to draw up a scheme of laws, 
which will at least copy the truth directly C00c), and to impose the 
severest penalties on their infringement. But he himself will act like a 
doctor who goes abroad,2 leaving written directions for his patients to 
1 300b. Contrast 296a-d. The resemblance between this teaching and that attributed to 
Protagoras in the Prot. is remarkable. See vol. m, 136-8. 
2 295 b-e. P. may have in mind himself and his colleagues, who visited several cities to draw 
up constitutions for them (vol. iv, 23 f.) and then left, though Campbell refers it in two different 
places to (a) gods who once looked after us but have now left us to ourselves (p. xlv) and (b) 
Solon (p. 141). More probably the going abroad refers only to the doctor in the simile. 
187 
Parmenides, Theaetetus, Sophist, Politicus 
follow. If he returns earlier than expected, and finds the condition of a 
patient changed, he will not hesitate to cancel his own earlier directions 
and take the case into his own hands again. So the statesman, who is his 
own legislator, will alter laws and customs as he thinks fit, his only 
criterion being the furtherance of justice and social benefit. 
As a tailpiece the known forms of constitution, excluding the 'true' 
one, are ordered according to the quality of life offered by each C00 c- 
303 b). Government may be by one man, by a few, or by the whole 
people (democracy). Only through laws strictly enforced can any of 
them emulate the one true state ruled by the one scientific or truly 
kingly statesman. Democracy, being the weakest, has least power either 
way, for good or evil, monarchy is best of all if under law (kingship), 
worst if lawless (tyranny), and in between come aristocracy and 
oligarchy, rule of the few under law and without it.1 But, he concludes, 
one can only wonder at the natural stamina of communities which 
survive under any of them, administered as they are by unscrupulous 
party-leaders with no understanding of the principles of government. 
(d) Final isolation of the statesman C03b~305 e). By a series oidiaireseis 
Plato has now marked off the statesman and his function from 
(i) others who might be called 'nurturers of mankind', such as 
farmers, bakers and traders B76 c); 
(ii) superhuman beings B74 c, 275 b-c); 
(iii) tyrants B76 c); 
(iv) producers of various material goods BSjb-S()b); 
(v) subordinate ranks in the community's hierarchy: servants, 
wage-earners and minor state functionaries or civil servants, including 
religious officials B89d-c>oe); 
(vi) bogus politicians, the statesman's closest imitators and rivals 
B9ia~303d). 
There remains one class especially difficult to sift out, for their  
functions are much closer to the statesman's, nor are they fraudulent  
impersonations of him but perform a valuable as well as a leading role in the 
community. Such are the masters of the arts of strategy or military 
1 This brief note may be instructively compared with the long description of the imperfect 
types of polity in Rep. 8 and 9, for which see vol. iv, 527-37. 
188 
Politicus 
command, of the administration of justice, and of oratory when it 
shares with kingship the task of persuading men to just ways.1 The 
problem is solved simply enough by yet another dichotomy of arts into 
an art itself—music, say, or a handicraft—and the superior art of  
knowing whether such an art is worth learning and how and when it should 
be employed.2 The commander's techne tells him how to conduct a 
campaign but not whether or not his country should go to war.3 The 
magistrate or juryman does not make the laws. His job is only to see 
that in any particular case they are administered without fear or favour. 
Oratory is in a similar position. All three, though genuine arts related to 
government and autonomous in their own spheres, are subordinate to 
the statesman's or kingly art which makes the supreme decisions. 
(e) The essence of statesmanship C05 c to the end). The aim of the  
discussion, we have been told, is first to isolate its subject, then 'to stamp 
upon it its unique form' (p. 167 above). We now enter the last stage. 
The form or essence of statesmanship is found in the art of 'kingly 
weaving', understood as combining disparate characters into the firm 
fabric of a stable community. When the bad material has been eradicated 
(p. 185 above), the acceptable citizens will fall into two main  
psychological groups, one marked by courage, boldness, enterprise and vigour, 
and the other by moderation, gentleness, the spirit of compromise. In 
private life, Plato notes, it is only amusing to see how the epithets 
change according to the natural affinities of the speaker C07 d): the 
'brave and energetic' become in the mouths of the others 'hard,  
insolent, manic', while they for their part convert 'gentle and moderate' 
into ' tardy, soft, cowardly'. But if either gains power their virtues may 
1 In Gorgias P. had condemned rhetoric outright. In Phaedrus he ridiculed contemporary 
rhetoricians but spoke of a * true oratory' based on knowledge. Here we have an intermediate 
class of those who, without possessing philosophical knowledge themselves, act under the 
instructions of the one who does and in the light of the 'true belief concerning what is just and 
good and their opposites' which he imparts to them C09 e-d) as the philosopher king imparts 
it to the subordinate guardians in the Republic. I hope this answers Skemp's question in n. i(iii) 
on p. 219. His description of the orator as a "'goverment spokesman'" is apt. 
2 When it is a question of defining the statesman P. of course reverts to the perfect embodiment 
of the art. He is not pronouncing on ethical problems arising from actual historical events, e.g. 
whether it was right for a doctor to obey a Nazi order to conduct experiments on human beings 
or for technologists to make the atom bomb possible but have no say in the decision to drop it. 
3 There is a parallel to this at Euthyd. 290c-d (noted by Skemp, 220 n. 1). 
189 
Parmenides, Theaetetus, Sophist, Politicus 
indeed turn to excess, and this is serious. One party, in its zeal for peace 
and quiet, may follow a policy of appeasement which may lay the 
country open to aggression by sapping its will to resist. The more 
militant sort may rush it into a series of rash ventures which by arousing 
the hostility of powerful neighbours lead to the same result. 
These are the main contrasting elements which the statesman must 
weave into his web as the firm, hard warp and softer woof. They will 
remain after the preliminary rejection of threads either too hard or too 
soft and weak to stand the strain, indeed both are necessary for the well- 
being of the state; but like the guardians of Rep. 2 they will be educable 
in the virtues of both courage and gentleness as distinct from their 
vicious excesses. They can be reconciled by bonds both spiritual (indeed 
divine, for men have a divine element in their souls) and natural or 
human. The first consists in education inculcating a true sense of values, 
a ' true belief soundly based on the knowledge which only the  
statesman is qualified to impart. This will purge the courageous type of any 
tendency to violence and turn gentle temperaments from a foolish  
softness towards moderation and good sense. Given generous natures to 
work on, the laws can foster a spirit of unity between virtues of  
opposite tendency. The more earthly bond is eugenic: check the tendency of 
similar types to marry, and encourage1 opposites to intermarry and so 
produce children with the virtues of both. This will be easy once the 
higher bond has been forged and all share the same standards and 
values. Then authority will be given to men of the mixed type, and 
when several act together the king will see that both types are  
represented, for the moderates are cautious, fair-minded and safe but  
deficient in drive, whereas the other set, though less careful and balanced, 
excel in getting things done.2 
So by his weaving the king unites individuals and the whole  
community, bond and free C11C4), in lasting concord and friendship, 'and 
1 There is no mention of legal enforcement as with the more drastic provision for community 
of wives among the guardians in Rep. (See 457 c.) Here P. speaks only of common ideals  
reinforced by honours, reputation and mutual pledges C10?). This advice on marriage recurs in 
Laws G73 a-d), where the idea of legal enforcement is explicitly rejected as productive of 
resentment and ridicule. 
* Anyone with experience on committees may question Plato's optimism about the possibility 
that anything will be accomplished by this carefully planned collaboration between progressives 
and conservatives. 
190 
Politicus 
in his oversight omits nothing conducive to such happiness as it befits a 
human society to enjoy'. 
E) Ethics and psychology 
The last few pages of the Politicus contain clear reminders of earlier 
dialogues, and have been seen by many as deliberately repudiating 
doctrines fundamental to the Protagoras and the Republic.1 The 
Protagoras is a full-dress defence of the Socratic thesis that the so-called 
different virtues are identical, to the extent that no one can possess one 
without the others and all alike are reducible to knowledge of good and 
evil. No 'part' of virtue therefore can conflict with another. In the 
Republic, as Skemp says B23 n. 1), all the virtues are harmonized and 
integrated in 'justice'. In the Politicus Plato goes out of his way to say 
that his present contention, that virtues can be in conflict, is something 
strange and surprising C06 a), though no ordinary person would have 
been surprised to learn that, as Protagoras maintained, a man might be 
brave but lack self-control or piety. Is this a signal that Plato is 
abandoning his own earlier teaching? According to Skemp, 'the new 
statement ... is equivalent to declaring eternal conflict between the 
warrior and civilian classes'. But the lesson of the new statement is 
precisely that the possessors of opposing virtues can and must be 
reconciled by wise government, and the Republic itself declares that' a 
gentle nature is opposed to a high-spirited ',2 but that the guardians must 
combine both. Naturally high-spirited or gentle temperaments are the 
raw materials of virtue on which the statesman-educator works, for 
both must be guided by reason (virtue, after all, is knowledge or  
wisdom) and the means to this is 'the one big thing', education,3 in 
Republic and Politicus alike. The Politicus repeats that they can be  
combined in one man, who should be given a position of authority, like the 
guardian. Harmony is achieved in both dialogues by recognizing that 
1 E.g. Skemp 222 n. 1 and 223 ?. ?: * the new statement must necessarily destroy the Republic 
psychological scheme'. Cf. Gomperz, GT 111, 184: 'a notable piece of self-criticism ... breach of 
Socratic intellectualism' (somewhat weakened by an uncalled-for reference to 'a breath of the 
Baconian, or modern inductive spirit'). Von Fritz {P. in Si^. 127) says P. departs from the 
Socratic principle enunciated in Prot. that you cannot have ??????? without ????????? or vice 
versa, but adds that an inclination to bold precipitancy (darauflosgeheri) or cautious holding back 
is expressly differentiated from the ?????? of ??????? and ?????????. 
1 ??????? ??? ???? ????????? ??????? ?????, 375 c· 3 Rep- 423e> 44ie—42a and elsewhere. 
191 
Parmenides, Theaetetus, Sophist, Politicus 
the state will contain contrasting psychological types and teaching them 
to live together. Neither the rash or violent, nor the foolish and weak, 
possess the virtues of courage or moderation, but aberrations from 
them. All excess is bad. Not for nothing has the Politicus given a homily 
on the importance of'right measure'.1 
I would therefore suggest, without dogmatizing, that if one forces the 
comparison, there is no direct contradiction between the Politicus and 
the Protagoras or Republic on the unity of the virtues or the scheme of 
psychology,2 although Plato speaks in simplified language of'virtues' 
as conflicting, instead of the more precise Qualities which when  
moderated and controlled by reason and education are parts of virtue'. There 
are of course differences, and Plato is not pressing us to take the two 
dialogues together. There is no one-for-one correspondence between 
the parts' of the soul and the two opposed temperaments, and the three 
social classes based on three psychological types are dropped. The  
careful education to ensure that reason predominates has been extended 
from guardians' to the whole citizen body, and so on. 
Conclusion 
This strange and fascinating dialogue is as much a work of art, or 
philosophical tapestry, as any other work of Plato's. He could have 
made his points that rulers must not be thought of as divine beings, or 
described as nurturers of a flock, or even the more general point that 
ours is an age of deterioration, without an elaborate cosmic myth; but 
he happens to like myths. He could have chosen another—perhaps more 
suitable—art than weaving to illustrate some principles of correct 
diairesisy but his fancy was caught by the idea of an inner affinity  
between weaving and statesmanship, just as in the Sophist he chuckled 
over the thought that the Sophist was first cousin to the angler. 
The Politicus does much to * bring philosophy', in the shape of the 
Republic, 'down from the skies'. Its central topics are the art of  
government at the human, Second-best' level and the correct application of 
the method of later dialectic. In spite of the myth, the other great sub- 
1 Already in Rep. D2365) ??????* is the epithet for the properly educated. 
* Cf. Tate in CR 1954, 116. Courage and ????????? are still 'parts' of virtue at 308b, and the 
idea of virtue and the virtues as a unity in diversity is still maintained in the Laws, at 963 c-d. 
192 
Politicus 
jects of the Republic—the relation between knowledge and true belief, 
the education of the philosopher-statesman in mathematics and dialectic 
culminating in the vision of the Good and the whole hierarchy of Forms 
beneath it, the nature of poetry, the divinity and immortality of the 
soul—get little or no mention. In its content the Politicus combines the 
ideal and the practical in a unique and puzzling way which probably 
reflects a transitional stage of indecision in Plato's own thought. Is it 
the ideal or the possible statesman that he has in mind in the final  
section on royal weaving? Presumably he is the one who has to use laws, 
and one remembers qualifying phrases like 'at least so far as possible in 
a polity' applied to the moderate character educated to a true belief, and 
4as much as it befits a society to be happy', at 309c and 311c. But apart 
from the language of perfection used elsewhere, are the preliminary 
purging and the divine and human bonds meant seriously? One cannot 
be sure,1 but the licence to kill, banish and enslave makes one think it is 
perhaps just as well that the true statesman has not yet appeared, and 
one may even hope that, if he should appear, that well-founded fear of 
the abuse of absolute power of which Plato speaks at 301 c-d may make 
people more chary than he expects of granting it to the superstatesman 
to inaugurate his reign of perfect felicity and righteousness. 
APPENDIX 
Elements of the myth in Plato and elsewhere2 
To judge by other examples, Plato in his myths took over much traditional 
material from earlier mythology and science, but built it into a new structure 
and drew his own moral. Here he openly acknowledges his debt to the stories 
of Atreus, the earth-born and the reign of Cronus, and mentions the services 
of Prometheus and Hephaestus. Apart from the reversal of cosmic motion, 
the following interrelated elements may be distinguished. 
1. The idea, familiar from many cultures, of the cyclic recurrence of 
historical events, credited in Greece to the Pythagoreans.3 
1 The elimination of bad material, or unfit stock, is discussed at greater length in Laws G3 5 a- 
36 c), where both drastic and milder measures are described. The subject is shelved, however, by 
the observation that if their theoretical plans for a Cretan colony should ever materialize, they 
could solve the problem simply by refusing entry and citizenship to unsuitable applicants. 
3 Skemp discusses the myth and its sources on pp. 82-108. 
3 See in general M. ?liade, Le my the de Veternel retour A949, Eng. tr. 1954). J. A. Philip's 
denial that Eudemus fr. 88 Wehrli shows it to have been Pythagorean seems to me wrong, in 
? 
I93 
GHG 
Parmenides, Theaetetus, Sophist, Politicus 
2. The Great, or Complete, Year, defined by Plato as the period required 
to bring sun, moon and planets back to the same relative positions ( Tim. 39 b), 
a matter, it was thought, of some 10,000 years, though theories of its  
conditions and duration varied considerably. For information about it see my 
In the Beginning 64 f. and 134 n. 2, and vol. 1, 282 f. and (in Heraclitus) 458 
with notes. Here it is strongly suggested by the description of the cosmic 
reversal as * the greatest and most complete of all tropai\ trope being the word 
for solstice, applied by Plato also to the planets {Tim. 39d). 
3. Connected with these was the belief in recurrent catastrophes by flood 
or fire {Tim. 22c), destructive of all civilization. Floods are linked by 
Aristotle to the notion of a Great Winter, corresponding on a far larger scale 
to the annual winter {Meteor. 352a28fF.), from which only a few ignorant 
hill-dwellers survived. In Plato these natural disasters reappear in the 
Timaeus B2 c) and Critias (i09d), and their effects are described in the Laws 
F77a-b, p. 330 below), and Aristotle takes seriously the belief that all 
knowledge and arts have been repeatedly lost and recovered. With the story 
of Deucalion (mentioned by Aristotle, I.e.), we may assume it to have been 
part of the common stock of Greek lore. For details see In the Beginning 
65-9. 
4. The age of Cronus, or golden age, was familiar from Hesiod and others. 
Empedocles had linked this with the cyclic theme in his doctrine of the 
alternate cosmic cycles, of increasing Love and increasing Strife, which put 
the golden age of innocence and happiness at the beginning of our own, when 
the power of Love was stronger. For this see vol. 11, 248 f. Besides the 
Politicus, the rule of Cronus occurs in Plato at Laws 713 a-b. There too 
Cronus appointed daimones to look after man. (The parallel with Pol. 271 d-e 
is close, and cf. Critias 109 b.) In general see Baldry's article * Who Invented 
the Golden Age?' in CQ 1952. 
5. The earth-born. In early Greek accounts, both mythological and 
rational, of the origin of life, the conception of the earth as literally the 
mother of the human race was widespread. It is fully treated in In the 
Beginning, chh. 1 and 2. For its place in the cosmic theory of Empedocles, 
see vol. 11, 206. Plato mentions the earth-born again in the * grand myth' of 
Rep. 3 (vol. iv, 462 f.), the Protagoras C2od), the Sophist B47 c) and the 
Critias (i09d). A variant occurs at Symp. 191c (men beget children  
themselves, but on the earth, not on women), reminiscent of the story of the birth 
of Erechthonius (Eur. fr. 925 N., Rose, Handbook no). 
6. The strange reversal of individual life, from age to infancy, occurs in a 
spite of the approval of de Vogel (JHS 1969, 164^). For the same idea in recent times see vol. 1, 
282 n. 1. 
194 
Politicus 
line of Hesiod, Erga 181, where he says that the present wretched 'age of 
iron' will be destroyed when men and women are born grey-headed. This, 
the only known reference before Plato, is not explicitly connected with a 
cosmic reversal, but its mention in the context of a succession of races, 
beginning with the golden one ruled by Cronus, does suggest that Plato has 
worked another old belief into his framework. Some have also seen it in 
Heraclitus fr. 88 DK: * Living and dead ... are the same, for the second 
changes and becomes the first, and the first, changing again, the second.' But 
see vol. I, 478 f. Theopompus, a younger contemporary of Plato, wrote of a 
tree whose fruit made men's lives flow backwards from old age through 
maturity and childhood to its end, and it is a matter of opinion whether he is 
more likely to have borrowed this from the Politicus or used a common 
source. (Theop. ap. Ael. VH 3.18, text in Frutiger, Mythes 243 n. 1. See 
Skemp in.) 
7. The miseries of primitive life and the benefits of progress. This 
repeats the account in the Protagoras C21 c-22c), which especially  
emphasizes the drawback of having no political art C22 b 5). The best pre-Platonic 
account is Aeschylus, ? V 442-68. 
Greek mythology and philosophy were divided between two theories of 
human development (described in detail in my In the Beginnings chh. 4 and 
5). One saw it as a degeneration from a primitive age of innocence as well as 
happiness, the other more rationally in terms of progress and improvement 
both technical and social, from wretched, animal-like beginnings. It looks as 
if Plato had ingeniously reconciled the contradiction by the conception of a 
universe undergoing periodic reversals of its motion, with contrary effects 
on the state of mankind. His nearest predecessor is Empedocles with his 
alternating eras of Love and Strife. But in Empedocles, first, the world is 
utterly destroyed at the end of both periods, which lead respectively to a 
complete, motionless fusion and a complete separation of the physical 
elements which by their partial mixture constitute a cosmos; no deity 
intervenes to save it at the last moment. Secondly there is no hint of the 
strangest feature of Plato's account, the reversal of cosmic revolutions. In 
this he has no Greek precursor, unless one regards as such the information 
acquired by Herodotus B.142) from Egyptian priests that during the 11,340 
years of Egyptian monarchy the sun reversed its course four times. The 
priests expressly said, however, that the changes brought no abnormalities or 
disasters on Egypt.1 This does not, any more than the Atreus portent, imply 
1 They also said that there were no gods in human form. I do not see how, as Skemp thinks 
(p. 91), this contradicts P.'s statement that a god, 'a different and superior being' {Pol. 271?), 
once looked after mankind. 
!95 
14-2 
Parmenides, Theaetetus, Sophist, Politicus 
Plato's extraordinary story of God turning the world for thousands of years 
and then * letting go the rudder' B72e) and leaving it to its own devices. 
Reitzenstein long ago suggested an oriental source, Zoroastrianism, but this 
has not stood up well to examination.1 The most natural conclusion is that 
this remarkable * single fact' behind the variety of traditional lore was the 
product of Plato's mythopoeic imagination. 
1 Reitzenstein and Schaeder, Stud. %. ant. Synkr. A926). For criticism see reff. in Leisegang, 
RE 2500 A941) and Koster, Mythe de P. etc., ch. vn A951). The point that P. expressly denies 
the theological dualism basic to Zoroastrianism B70 a), though not overlooked by R. (see 
Koster 42), certainly did not strengthen his case. See also Skemp 92-5 and 146 n. 1. 
196 
Ill 
PHILEBUS1 
We are speaking to men, not gods, and the most natural concerns of 
human kind are pleasures, pains and desires. 
Plato, Laws 73 2 e 
Date.2 The Philebus is generally held, on grounds of both style and 
content, to be one of Plato's latest dialogues. Bury A897, p. ix) could 
call this 'the verdict of the most recent critics', and it is still that of the 
majority, e.g. Crombie A962): 'It is certainly one of the latest group of 
dialogues.' Ryle however A967) placed it in the middle group, though 
later than the Timaeus for which he adopts Owen's dating. Some close 
parallels between the two dialogues strongly suggest proximity in time, 
of which anyone attempting to date either must take account, whether 
he believes, with Ryle, that Philebus ' echoes' Timaeus, or with Leise- 
gang, Dies and Hackforth that it leads up to it.3 it will be treated here 
as a late work, following on those in the previous chapter. 
Characters. Socrates again leads the debate, probably because the  
subject is pleasure, about which he has so often expressed strong views in 
earlier dialogues, but his manner is more like that of the Eleatic visitor 
than of the ironic Socrates we know. To lend credibility to this, Plato 
makes Protarchus, in the name of all the young men present, demand 
1 Besides the published sources mentioned, I wish to acknowledge a debt to Cornford's MS 
notes for a lecture-course on the Philebus. Gosling's Philebus A975) appeared after this chapter 
was written, but I have added a few references to it. 
3 The authenticity of the dialogue is no longer in doubt. In the last century Schaarschmid and 
Horn denied it on the ground that it ignored the theory of Forms and was in other respects 
unplatonic. See Bury's ed., p. lxiv, and for a point-by-point refutation of Horn, Rodier's chapter 
on the Phil, in his Etudes. 
3 Bury, ed. p. ix, Crombie, EPD 1, 252, Ryle in Ency. Phil, vi, 320 and P.'s P. 251 and 285, 
Leisegang in RE 2505, Dies, Bude ed. cii, Hackforth, PEP 3. In putting Phil, after Tim. Ryle 
follows Wilamowitz, PI. 1, 628, 635. To put it nevertheless in an earlier group is to revive a still 
older view. For details see Bury lxxxvii-lxxxi. It should be added that Owen, though assigning 
Tim. to the middle group, still regards Phil, as late and as representing a changed Platonism. See 
his 1953 article reprinted in SPM> 315, 321, 324 and (especially) the final paragraph on 338. 
197 
Philebus 
that he abandon his practice of getting everyone into difficulties by 
asking unanswerable questions. This time he must resolve their  
difficulties, and they will follow him as best they can. (i9e-2oa. Cf. 28b-c.) It 
is one of few personal touches in a dialogue on the whole lacking in 
dramatic interest.1 
The name Philebus is unknown to history. Some have thought it 
invented: he is simply the embodiment of a dogmatic hedonism.2 This 
would certainly fit his role. After being introduced as maintaining that 
gladness, pleasure, enjoyment and all that go with them constitute the 
good for all living creatures ',3 he turns his back on the discussion named 
after him, and leaves it entirely in the hands of Protarchus. 
Protarchus is called son of Callias and a follower of Gorgias A9b, 
58a), and could be the man mentioned in Aristotle's Physics, 197b 10. 
This suggests a historical figure, though some think him imaginary. 
Callias was a common name, but there is no reason why he should not 
have been one of the two sons of the well-known Sophist-fancier (vol. 
iv, 216).4 Though calling himself a hedonist like Philebus, he is a very 
different character. He may stick to his guns on a point like the  
nonexistence of false pleasures (p. 218 below), but agrees without argument 
to some fundamental Socratic theses, such as the government of the 
universe by divine Reason B8 d-e), which make his ultimate conversion 
certain. He is young A6b). 
1 Nevertheless Hackforth {PEP 7) mentions a number of places where S.'s familiar traits come 
through. Wilamowitz A, 628) took the quantum mutatus view: S. has turned into a 'thesis- 
defender'. 
3 Wilamowitz 1, 629, Friedlander in, 309, Leisegang, RE 2502, Hackforth p. 6 with n. 3. As 
to his name (???-????), S. is surely witness that a ' lover of youth' need not necessarily be a 
champion of extreme hedonism. Nor need a ????? be an ???????. Gosling's 'Loveboy' is  
ambiguous: cf. 'playboy'. 
3 Philebus has maintained not only that 'good' is rightly predicated of pleasure (as 11 b alone 
might suggest), but that 'pleasant' and 'good' have the same denotation (are properly used of 
one single ?????: see 60a). 
4 Taylor (PMW 409) thought it impossible because ApoL 20 a represents them as boys at the 
time of S.'s trial; but S. specifies no date for his conversation with Callias about their education. 
Among previous scholars Wilamowitz A, 629) and Friedlander (in, 310) thought Protarchus 
historical, Hackforth (p. 7) imaginary. 
I98 
Philebus 
Pleasure and good. Since Plato's subject is the place of pleasure in the 
good life, something must first be said about the two key-concepts, of 
pleasure and 'the good',1 as he uses them in this dialogue. 
Pleasure. Plato nowhere defines pleasure, and indeed the word's field 
of reference changes during the discussion. The Greek word ''hedone\ 
like the English pleasure', was applied widely.2 A glance at the lexicon 
shows that it was not confined to sensual pleasures or what are called 
here the pleasures of Replenishment', the sort to which, as in the 
Protagoras, one 'yields', which one 'cannot resist', which 'defeat' one. 
Yet this was the most popular usage, which Plato adopted without 
examination when in the Gorgias he wanted to combat hedonism in its 
most extreme form. According to its champion Callicles, pleasure 
attends the process of satisfying wants like hunger and thirst: with their 
satisfaction the pleasure ceases, and since pleasure and good are identical 
—there are no bad pleasures—happiness lies in letting the appetites 
grow as big as possible and ensuring the means of feeding them D94a- 
95a). It is from this simple conception of pleasure that the Philebus 
starts,3 when it is sharply opposed to intellectual activity as its rival for 
the title of'good'. As the discussion proceeds, however, we find that 
one of Plato's main aims is precisely to analyse the concept of pleasure 
as a necessary preliminary to estimating its value in human life. 
Hedonism in various forms was being actively preached by Plato's 
acquaintances, including Aristippus representing the extreme, Calliclean 
type, and Eudoxus who, says Aristotle, combined it with a life of 
unusual temperance and self-control.4 It badly needed the attentions of 
1 I have omitted pleasure's rival, intellect, as to which only a reminder should be needed of 
how Plato associates various terms covering thinking, the mind, and knowledge (????????, voOs, 
???????? and others) more closely than we do. ???????? is at one moment what is known, the 
content of knowledge including practical crafts, and at another equated with voOs, the mind that 
knows. As opposed to pleasure it must be thought of as mental activity, yet in the classification 
of what is called voOs ??? ???????? at 55cfT. all the emphasis is on the subjects known, from 
husbandry and building to pure mathematics. If the Greek assimilation of these terms is not 
familiar by now, see vol. in, 501 n. 3 and iv, 265. 
2 Nor is the discussion confined to this word, but ?? ?????? ? and ?????? are thrown in as 
working equivalents (nb). Cf. S.'s depreciation of Prodicean niceties at Prot. 358a. Thus 
pleasure and enjoyment are not separated in Plato as they are, for instance, by a modern 
philosopher like Bernard Williams in PAS suppl. vol. xxxm, 67. 
3 As an argument ad hominem this is fair enough, for it is Philebus's conception of pleasure. 
4 Arist. EN 1172b9-i6. For Eudoxus see pp. 453-5 below, for Aristippus, vol. in, 490-9. On 
the relation of P.'s discussion of pleasure to current debates in the Academy see Jowett's editors, 
vol. in, 532f., n. 1. 
199 
Philebus 
a skilled dialectician to clear up the confusion caused by the use of a 
single term to cover a complex of different, even incompatible  
experiences. Pleasure, Plato claims, may be 'true' or 'false', and 'true' or 
4pure' pleasures turn out in the end to be closely related to wisdom and 
knowledge F3 e). In this way pleasure as such can before the analysis be 
characterized as apeiron, without bounds or measure B7 c, 31a), whereas 
later 'pure' (opposed now to intense') pleasures are emmetray bound 
by measure E2c); and again at 65 d pleasure (glossed as excessive  
enjoyment to show that the word is used in the old sense) has relapsed into 
the measureless. The crude question: 'Is pleasure good or bad?' is 
unreal until one has answered the further questions: 'What sort of 
pleasure?' and ' Pleasure in what?' This is a big advance on the Gorgias 
and Protagoras and even on the more elaborate treatment of pleasure in 
Republic 9; and it owes much to that 'gift of the gods' A6c), the 
dialectical art of collection and division evolved, from a hint of Socrates,1 
in the other dialogues of Plato's late maturity.2 
The good. In defending the identity of pleasure with goodness3 before 
the dialogue began, Philebus has said that by 'good' he meant what is 
good for every living thing A1 a, 60a), as in real life Eudoxus had 
included 'creatures both with and without reason' (Arist. EN 1172b 10). 
Socrates on the other hand, by his original claim that it resides in 
intellectual activity, or as he modifies it in the dialogue, that thought is 
the superior element in the mixed life which is the best, has immediately 
confined it to the human race.4 'Pleasure must be the good because 
every creature pursues it', said Eudoxus. 'Pleasure is not the supreme 
good, though every ox and horse and other beast proclaim it by their 
1 See p. 27 above. 
3 A. Hermann in Untersuchungen \u P.'s Auffassung der Hedone pursues the conception of 
pleasure through the dialogues and concludes on p. 77 that they show 'a planned and systematic 
development'. The obscurities of Phil, are reduced if one can see it historically, taking into 
account the contemporary controversies in which P. was involved. If this chapter has not said 
enough about them, Thompson's lecture 'Introd. Remarks on the Phil.\ printed in/, of Philol. 
1882, should still be read for its information on the Cyrenaics, Cynics, Megarics and  
Pythagoreans. 
3 That ?????? at nb means not only 'good' but ''the good' is clear from the context and 
confirmed by the substitution of ??????? in the recapitulation at 60 a. 
4 The necessity of thought must clearly override the few passages in which, whether from 
indifference to detail or a willingness to humour the hedonists where the distinction was less 
important, S. uses their expressions. See 60 c, 22 b. At 22b the addition of ?????? is remarkable, 
and was arbitrarily excised by Badham. 
200 
Philebus 
behaviour' is Plato's reply F7b). Thought and its congeners—memory, 
right belief, true reckoning—are better than pleasure for all capable of 
engaging in them A1 b). What they seek is 'a habit {hexis) and  
disposition of the soul capable of making the life of all men happy', elsewhere 
expressed as ' the best of human possessions' A1 d, 19c); and it amounts 
to asking the old Socratic question posed in the Gorgias: 'How ought 
one to live?' 
In the Republic too E05d) the good was described as 'what every 
soul pursues, for the sake of which all its actions are performed', but the 
Philebus offers not the slightest hint of a culmination in any mysterious 
Form of the Good, transcending knowledge, truth and even existence. 
What it offers is far less open to Aristotle's criticism of'a Good solitary 
and apart, which even if it exists can clearly never be practised or 
possessed by man' {EN 1096b32). Rather does it resemble Aristotle's 
own teaching, that human excellence {arete) is a hexis, and the good for 
man (which both philosophers identified with eudaimonia, happiness) 
an activity of soul dependent on that hexis {EN no6b36, 1098a 16). 
Here again we notice that tendency away from the ideal to the  
practicable, the humanly attainable, which marked the Politicus. It is a fine 
thing to know the divine Form of Circle, but it will not help us to find 
the way home F2b). As we shall see later, this by no means implies 
cutting off humanity from what is above it, the eternal Reason which 
controls the cosmos, and of which man's own reasoning powers are a 
part. Nor need it involve sacrificing the belief in transcendent Forms, 
even if they no longer occupy the centre of the stage (p. 237 below). 
Three criteria are mentioned by which any candidate for the title 
'the good' must be judged: it must be perfect (complete in itself, 
fulfilled),1 adequate (so that whoever possesses it needs nothing  
besides) and the universal object of choice. (See 20 d, 22 b, 60 c, 61 a.) The 
nearest to a definition of it comes near the end, at 65 a, where Socrates 
says that it combines three forms: beauty, symmetry or proportion, and 
truth or reality. These, they agree, are furnished by mind {nous) rather 
than by pleasure. Nevertheless mind by itself is not in human life either 
1 ??????, having achieved its telos, i.e. not only full internal development but purpose or aim. 
Similarly its contradictory ?????? meant not only incomplete or imperfect but ineffectual, a 
failure. 
201 
Philebus 
the sole or the supreme good. It has already been agreed that no one 
would choose either pleasure or intellectual activity alone Bia-e, 
6od-e), and in the final five-fold classification of human possessions', 
intellect and thought are placed third in order of merit. Further  
comment on this must come later. 
Subject and scope. Most commentators at some stage emphasize the  
obscurity of the Philebus.1 Plato's principle of going ' wherever the wind 
of the argument blows' leads to some disconcerting transitions, and the 
doctrines themselves, while reminding us of those in other dialogues, 
give them at the same time a new twist and are bafflingly complex. This 
chapter cannot hope to solve all the problems, but will try to set them 
out fairly and assess the various solutions which have been proposed. 
Let us first of all be clear what is the subject under discussion. It is 
not what Taylor said it was (PM/F408), ' a straightforward discussion 
of whether the "good for man" can be identified either with pleasure 
or with the life of thought'. That has been argued out (evidently to a 
stalemate) before the dialogue began. The dialogue opens when 
Philebus, exasperated perhaps like Callicles by Socrates's addling little 
questions', hands over his part to Protarchus. After a brief  
recapitulation Socrates radically changes his ground with the suggestion that 
perhaps neither pleasure nor thought in isolation can provide a happy 
life. Both have their place, and the question he wishes to discuss with 
Protarchus concerns their relative importance. First prize for goodness 
must go to the ' mixed' life, and what remains to be decided is whether 
pleasure or thought shall have the second. By this dramatic device of a 
dialogue before the dialogue Plato shows plainly that he has no  
intention of treating us to yet another refutation of the naive hedonistic 
equation of pleasure with good which he had already dealt with amply 
in the Gorgias and Republic. With the question 'what place can be 
assigned to pleasure in the good life, and what sorts of pleasure can 
there find admission?' he breaks new ground.2 This explains both his 
1 A good example is Grote: 'It is neither clear, nor orderly . . . Every commentator of Plato, 
from Galen downwards, has complained of the obscurity of the Philebus.' This 'remains 
incorrigible*. (P/. n, 584 with n. u.) 
3 The above owes much to some remarks of Hackforth's on p. 112 of PEP, from which I 
have quoted the question above. Among previous discussions of pleasure, perhaps that in 
202 
Philehus 
reason for writing another dialogue about pleasure and the complexity 
of its argument. 
One must also note at the outset a premise on which Plato's whole 
case rests, that is, his conviction that the world is rationally ordered by a 
divine Intelligence which, as he argued in the Phaedo (97c), naturally 
aims at the best. Its effect has been to introduce order, harmony and 
measure, and through them beauty and goodness. This runs right 
through his thought from Phaedo and Republic to Laws, and is the main 
theme of the Timaeus. Its evidence, he thought, is daily before our eyes 
in the order, beauty and regularity of motion of the sun, moon and 
stars,1 governed by mathematical ratios and giving birth to time, the 
uniform succession of days, nights and seasons on which the life of all 
creatures depends. Thus Plato's exalted view of measure, proportion, 
symmetry and so forth is governed throughout by a universal  
ideological hypothesis: what makes any mixture good is, as in the macrocosm, a 
due measure and proportion among its ingredients. He is emphatic 
about this. (See 64d-e.) The Philehus is an excellent illustration of 
Plato's talent for combining the ethical and the metaphysical, the human 
and the cosmic.2 The whole of reality is his province, and he is  
unwilling to separate any of its parts since for him they are parts of an organic 
whole. Man's soul is a fragment of the universal soul C0 a), order is the 
same in individual souls, in the city-state and in the universe at large. 
The Philehus treats of it in the individual, the Politicus in the state, and 
the Timaeus in the universe at large, but all alike are at pains to put 
mankind in his setting as an integral part of the cosmic order. 
The argument. A brief outline may help to keep the connexion of 
thought in mind when we turn to details. 
In pitting pleasure against thought as 'the good', Socrates and 
Philebus may have been wrong. What if it is some third thing, with 
pleasure and thought competing only for second place? Both are genera 
Protagoras comes nearest to Philebus. Whatever view one takes of the seriousness of S.'s  
hedonistic thesis in that dialogue, the 'pleasure' which he advocates contains a strong admixture of 
thought in its 'art of measurement' and canny foresight. 
1 Mentioned at Phil. 28 c For the divine mind see, besides 28c-e, 22 c and 3oa-d. 
2 For the Victorian rationalist, George Grote, 'the forced conjunction of Kosmology and 
Ethics' was 'the one main defect' pervading the dialogue {PI. 11, 611). 
203 
Philebus 
containing dissimilar species, certain pleasures may be good, others bad 
(though at this stage Protarchus maintains that any pleasure, qua 
pleasure, must be good). 
But this raises the old question of one and many, how a single Form 
can exist, retain its unity, and enter into its many transient instances. 
To answer this Socrates describes a method resembling the dialectic, 
or division into kinds, of Sophist and Politicus.1 Protarchus and Philebus 
doubt its relevance, but if Socrates proposes to investigate the different 
kinds of pleasure and wisdom, well and good. He replies that the task 
may be unnecessary, because neither pleasure nor thought by itself 
satisfies their agreed criteria of goodness, which demand a ' mixed' life 
containing both. No one would wish for a life of pleasure devoid of all 
thought, memory and kindred faculties, nor one of cerebration with no 
pleasures. Yet either pleasure or wisdom may win second prize by 
being the element which makes the mixed life good, and this he claims 
for wisdom. 
In all that exists four forms, or kinds, can be distinguished: the 
unlimited, limit, their mixture and the cause of the mixture,  
unlimited ' designates what admits indefinitely of more and less, like size 
or speed, * limited' includes proportionate relationships expressible 
numerically (half, double etc.), their mixture results in goodness and 
harmony, and the cause, or maker of the mixture, is reason, supremely 
manifested in the divine Mind that regulates the universe. Pleasure 
belongs to the unlimited because, as its advocates themselves say, it 
always admits of more. But that cannot be a reason for its goodness, for 
it is equally true of pain. 
Pleasure in fact cannot be considered apart from pain. Pain occurs 
when the internal harmony of a living creature (belonging to the mixed 
class) is disturbed, and pleasure accompanies the process of return to its 
natural condition. That is one kind,2 but there are also pleasures of 
anticipation, i.e. mental pleasures. But anticipation may be disappointed, 
and its pleasure dependent on a false belief, in which case Socrates 
maintains that the pleasure itself is false, which Protarchus stoutly 
denies: beliefs can be true or false, but not pleasures. In a long discus- 
1 Cf. 23C4 ???? ??????????, 12 ????, ??. ???' ???? ???????. 
2 So pleasure is to be analysed after all. 
204 
Philebus 
sion, Socrates suggests other ways in which a pleasure may be false. 
First, as with vision, in judging pleasures from a distance one may be 
deceived about their magnitude (intensity), and the amount by which a 
present pleasure exceeds or is exceeded by the pleasure of anticipation 
may be called false pleasure. Secondly, one may confuse absence of pain 
with positive pleasure, in which case the impression of pleasure is false. 
This reminds Socrates of thinkers who claim there is no such thing 
as pleasure, but only escape from pain. They go too far, but their 
reminder that many pleasures are not pure pleasures, but mixed with 
pain (as drinking with thirst), is salutary. Yet pure pleasures there 
undoubtedly are. They include appreciation of form and colour, many 
sounds, and even scents; and lastly the pleasures of learning. None of 
these is preceded or accompanied by any painful sense of want. Further, 
4unlimited' applies only to the mixed pleasures going with bodily 
replenishment; the pure should be classed with limit or the measured; 
and the truth (reality) of pleasures is decided by their purity. 
Pleasure is a process, not a finished state, or product, and processes 
are means not ends, as shipbuilding is for the sake of the ships.  
Therefore not pleasure, but a higher end, should be called 'good'. Is it not 
unreasonable also to admit that the good resides in soul rather than body 
yet confine it to pleasure and deny the title 'good' to wisdom, courage 
and other virtues? Is one to equate the suffering of pain with lack of 
virtue?1 
Now it is the turn of wisdom and knowledge to be analysed and 
tested. Knowledge is divided into practical, or technical, and cultural, 
and the former judged by the extent to which it uses exact canons of 
measure and number, that is, approaches pure knowledge. The study 
of number is itself twofold, philosophical, operating with abstract, 
equal units, and popular, whose units are unequal—encampments, cows, 
anything whether large or small. The purest, truest knowledge is 
dialectic, whose objects are the Really real', always the same,  
unchanged and unmixed', and their closest kin. 
Both ingredients of the best life having been analysed separately, it 
1 Even in the loose construction of Phil, these arguments are exceptionally isolated from any 
context, and one would hope too that, as Hackforth says {PEP 111), they were not intended to 
stand on their own feet as a serious refutation of hedonism. 
205 
Philebus 
remains to consider how to mix them, and in particular which species of 
each should be included. As to knowledge, to live a human life at all one 
cannot confine it to dialectical philosophy. Not only the Forms, but 
their imperfect embodiments must be studied, even those arts which 
rely on empirical guesswork rather than calculation. Of pleasures, the 
true will be admitted, and any others that may be either necessary  
(presumably as attending the satisfaction of basic physical needs) or  
compatible with health, temperance and the other virtues. But Intelligence 
itself will reject the intense pleasures which go with folly and vice, for 
the simple reason that they would make its own operations impossible. 
Finally there is a third ingredient to add, namely truth or reality, for 
the best life must be capable of realization. 
The good for man, then, does not lie in either pleasure or intellectual 
life alone, but is a blend of both, containing all kinds of knowledge, and 
the better kinds of pleasure. One last question remains: What is the 
most valuable element in the mixture, that which makes it so desirable? 
When they know this, they can consider whether pleasure or knowledge 
comes closest to it.1 But what makes any mixture good is its formula, 
the proportion in which its ingredients are blended. It is due measure 
that imparts goodness, and beauty and truth as well; and Protarchus 
needs no persuading that the achievement of measure and proportion is 
due to rational activity rather than pleasure. So intellect gets second 
prize. The dialogue ends with a list of five ' human possessions' drawn 
up in order of merit. 
The One-and-Many Problem (i4c-i6a). At 12c Socrates declares that 
though pleasure is one thing it takes many forms, unlike and even 
opposed to each other, as black and white are contraries though both 
embraced by the same genus, colour; so one pleasure may be good and 
another bad. By allowing that his own favourite, wisdom, is in the same 
case, he gets Protarchus to withdraw his objection that though pleasures 
may arise from different causes, in themselves they cannot be opposed. 
Protarchus perhaps gave up too easily. Socrates is treating pleasure 
as a genus: there are different kinds of pleasure. In illustration he cites 
1 So in spite of intervening turns and twists, the original plan has been faithfully executed. 
See nd-e, and the careful recapitulation at 666-67 a. 
206 
Philebus 
the pleasures of the licentious and the temperate, the foolish and the 
wise. This, as Protarchus says, is to differentiate them by their objects, 
by what induces the pleasures. He himself understands by pleasure the 
subjective feeling alone. The gluttony or lechery, the knockabout farce 
or the improving book which give pleasure to different people are not a 
part of the pleasure itself. This was the view of the Cyrenaics.  
According to them ' pleasure does not differ from pleasure ... Pleasure is good 
even if it arise from the most unseemly sources ... Even if the action be 
unacceptable, the pleasure by itself is choiceworthy and good' (D.L. 
2.87-8). This is relevant to the later question of true and false pleasures, 
and it would have been interesting to have it thrashed out. 
They are now faced, says Socrates, with the whole troublesome  
question of how one thing can also be many, not as it applies to physical 
objects, either as wholes of parts or in their mutual relations (one thing 
both large and small etc.), but as it arises when one posits single Forms 
like Man, Ox, the Beautiful or the Good. The points of doubt are three: 
A) Should one accept such monads as really existing (????? ?????)? B) As 
an immutable and eternal unity, how can each Form be this one thing [i.e. 
contain both Unity and Being]? C) What is its relation to the infinite 
multitude of things that come to be [sc. in its image]? Does it become many 
by being distributed among them or does it achieve the apparent  
impossibility of getting apart from itself and appearing as a whole in one and many 
at the same time?1 
The question of one thing being both large and small was mentioned in 
the Republic E24a-b) as a useful propaedeutic and stimulus to thought, 
and dealt with in the Phaedo by the concept of participation A02b- 
103 a). More generally, to puzzle over one thing having many attributes 
was dismissed in the Sophist2 as due to poverty of intellect'. More 
relevantly to the present passage, participation is the first solution to the 
one-and-many problem offered, and subsequently criticized, in the 
Parmenidesy where also all the three Serious' questions raised here were 
debated.3 
1 15 b. Some have seen only two questions here. But ?????? ??? . . . ???? ?? . . . ???? ?? 
????? makes P.'s intentions clear, and the three are not hard to distinguish. I am not convinced 
by Striker's arguments (Peras u. Apeiron 14 n.) for a different interpretation of ???? ?? ????? in 
this context. Crombie (EPD 11, 362 n.) resorts to emendation. 
3 251 a-b; see pp. i48f. above. 3 See pp. 36f., 40 f., 54 above. 
207 
Philebus 
It is tempting to translate the questions into modern terms as asking 
how one subject can have many predicates (trivial) and how one  
predicate can have many subjects, or what is the relation between a universal 
and its particulars (serious). But these would be different questions. The 
Serious' questions would today be treated purely as questions of logic, 
but for Plato the so-called predicate', or 'universal', existed \n a  
supersensible world. It is not a question of logic but of the structure of 
reality. So his first question is: Do such monads as the Forms exist? 
The second is Parmenidean: Since Unity and Existence are two, not 
one, how can a unit both exist and maintain its unity? It would already 
be a combination of more than one Form, as the Sophist says.1 The 
third was raised at Farm. 1306-31 e and never directly answered. 
A brief addition, difficult to translate,2 shows that in Plato's mind the 
question is closely bound up with speech and its mental counterpart.3 
The enquiry, it says, must start from the admission that one and many 
become the same through discourse (???? ?????). Their identity pervades 
everything that is said, is a permanent feature of speech and thought and 
something inherent in our own nature, and a young man's first  
awareness of it is an intoxicating experience. From this point the argument 
goes on to show, first, that our understanding must be disciplined by 
dialectic, and second, that, so disciplined, the form of our sentences and 
thoughts does reflect the structure of reality and is not only something 
'in us', i.e. subjective. 
Dialectical solution of the one-and-many problem (i6b-i8d).4 There is a 
method 'easy enough to indicate but hard indeed to practise', to which 
Socrates is devoted though it has often left him deserted and helpless,5 
a veritable gift of the gods, to which, he claims A6 c 2-3), is owed every 
1 Parm. 142b—c, p. 54 above. Cf. Soph. 250a: If Motion and Rest exist, there must be ?????? 
?? ???? ?????, ?? ??. 
2 Cf. the renderings of Hackforth (p. 22), Bury (xxxv) and Ackrill, introd. to Stenzel's PMD, 
xxxi. 
3 ????? covers both speech and thought, conceived of as an internal dialogue of the mind with 
itself (Soph. 263 e). 
4 Anyone who reads the following pages on dialectic and the Fourfold Analysis must be 
directed also to Crombie's long and meticulously careful discussion of them in EPD 11, 359-70 
and 422-40, which differs on several important points. 
5 P.'s insistence on the extreme difficulty of success in the dialectical method is noteworthy. 
Cf. p. 167 above. 16b6 ???????, as at Phdr. 266b. 
208 
Philebus 
discovery of art and science. It is based on the truth that 'all things that 
are ever said to be consist of one and many, and combine in their nature 
limit and unlimitedness'.1 So (i6cio), 
Things being thus ordered, we must assume a single form for every thing and 
look for it—for we shall find one there—then after one form two, if there are 
two, or else three or some other number; then do the same with each of these 
units until we discover that the original unitary form was not simply one, 
many, and an infinite number, but how many. We shall not introduce infinity 
into the plurality until we have seen the whole number which lies between 
infinity and the one. Only then shall we release each unity in every thing into 
the infinite . . . But our clever men of today posit their one arbitrarily and 
their many too quickly or too slowly.2 They leap straight from the one to the 
infinite multitude, and the intermediates escape them. This is the difference 
between eristic and dialectical discussion.3 
This is a more elaborate and obscure description than those in the 
Sophist and Politicus (pp. 129^, 166 above), but the upshot seems to 
be roughly similar. With them also in mind, we may say that any group 
of phenomena to which we rightly give the same name will be found to 
have a common nature or form.4 With 'every thing' (???? ??????) at 
i6d2 I take it Plato has chiefly in mind sensible particulars, also called 
'the unlimited'. Even the philosopher or scientist must start from 
these,5 though scientific thinking is concerned with the higher levels.6 
1 Peras (limit) and apeiron (infinite, indefinite or unlimited in number, quantity or degree). 
For their meanings and strong Pythagorean associations (well known to Plato) see index to vol. i, 
s.vv. and p. 532. (Gosling uses 'determinant' and 'indeterminacy'. See his note on i6cio, p. 84.) 
It may be confusing, but must be accepted, that P. here uses apeiron numerically, for the  
uncountable multitude of particulars in a species, and in the Fourfold Classification qualitatively, to 
signify the indefinite possibilities of variation in temperature, strength, speed and so on. (For a 
rather different possibility, see Gosling, Phil, xiii, xvii.) 
2 ?????????? has been found strange, because P. has only mentioned the fault of jumping too 
quickly from a genus straight to particulars. But it is an equally possible fault in classification, 
which he presumably mentions for the sake of completeness. One may wrongly interpose what 
are not true species, because their differences are non-essential, as for instance with biological 
'sports'. See also Rodier, Etudes 76f., and for Gosling's interpretation his Phil., p. 85. H. Maier, 
Syll. des Arist. 2. Teil, 2. Halfte, pp. 5 f., interpreted this sentence as an attack on the Megarians. 
3 The same distinction between eristics and philosophers goes back to the Rep. See 454a. 
4 See Rep. 596a and vol. iv, 550. Names depend on forms or essences, not vice versa (pp. 27- 
29 above). 
s Cf. 18a ???? tis ?? ??????? ????????? ?????? ?????????. 
6 This perhaps needs a little enlargement, since it has worried commentators. Neither P. nor 
Aristotle differentiates between the relation of genus to species and that of species to individual. 
It is only that beyond a certain point philosophical knowledge cannot penetrate. It must go as 
15 
209 
GHG 
Philebus 
The first step, then, is to identify a generic form in a multitude of 
instances;1 then to work downwards, dividing it into species, and those 
into sub-species, until the lowest definable class is reached. Beyond that 
the philosopher cannot penetrate. There remains only the unlimited and 
not further definable mass of individuals from which we started. Plato 
would have agreed with Aristotle that ' there is no definition of them 
but they are known by perception and intuition' {Metaph. 103635). 
Nevertheless, though we perceive the individual, perception is in a 
sense of the universal: seeing Callias gives us our first impression of the 
species man, to which Callias belongs {An. Post. 100 a 16). There is this 
unique human faculty, described by Plato at Phdr. 249 b, of advancing 
'from many sensations to a unity pulled together by reason'. 
Socrates follows this description with two2 illustrations, of which the 
second at least does not seem to exemplify the method of genus-species 
division,3 However, we shall shortly see it applied to the concept of 
pleasure, for the sake of which it was introduced. People experience on 
innumerable occasions feelings which they deem sufficiently alike to be 
given a common name, 'pleasure'. It is 'one in genus', but its 'parts' 
differ widely.4 But it cannot be evaluated by saying simply that it is a 
unity with many parts and innumerable instances: the species must be 
far as it can, e.g. in 'dividing' animals one must not stop at dogs, but continue to divide into 
Alsatians, spaniels, fox-terriers etc. These still present specifiable, or intellectually separable, 
differentiae (in Platonic-Aristotelian language, they have a comprehensible eidos), but the two 
King Charles spaniels Fido and Bruno no longer do so, though the senses can tell one from the 
other. At this point one must, for scientific purposes, let things 'slip away into the apeiron'. For 
Aristotle this was philosophy's greatest crux, for he had to reconcile two theses, both of which he 
believed to be profoundly true: (i) The philosopher's task was to explain substance or reality 
(tis ? ?????, Metaph. 1028 b 4), B) the primary realities are individuals. For his statement of the 
dilemma see Metaph. 999324-9. But this is not the place to pursue Aristotle's full solution. 
1 P. says nothing here about definition, but what we have is the first stage described at Phdr. 
265d as 'bringing the dispersed plurality synoptically under one form, in order to define . . .'. 
P. never gives as much attention to collection as to the subsequent division, but it is mentioned 
by name (????????) at 25 d. 
2 Generally reckoned as three, but see below. 
3 See Hackforth 24f. and Ackrill in Ryle, 380. Trevaskis in Phron. i960 has made out an 
interesting case, based on the illustrations, for holding that the method described is not that of 
division at all. But one may agree with Runciman (PLE61 n. 2) that 'it is perhaps more  
appropriate to see it as a further development of the method of diairesis'. 
4 12c For 'parts' = species cf. Pol. 262b, Euthyphro i2c-d, and p. 153 n. 4 above. All 
members of the same genus must have at least one feature in common (or so it appeared to Plato 
and Aristotle), even though in other respects opposed. In S.'s example of colours, black and 
white and all the rest possess visibility, as all animals resemble each other in being alive, though 
as different in other respects as tiger and mouse. 
2IO 
Philebus 
enumerated and defined. We are indeed spared the successive  
dichotomies of the Sophist and Politicus, and the classification may not seem 
so methodical, since by the adoption of different principles of division1 
some classes overlap or even coincide with others, as pure with true and 
mixed with false. The outcome, however, is clear and instructive as 
Socrates proceeds to distinguish pleasures of replenishment, of  
anticipation, of body, soul, and both together, pure pleasures and mixed, true 
and false, and subdivides false pleasures into three. Dialectical method 
is no bed of Procrustes, forcing every subject into a single rigid  
framework. Its general rules leave plenty of scope for philosophical initiative, 
which is why it is so difficult. 
Note on the * third illustration9. At 18a Socrates says he will add one further 
point before returning to the subject of pleasure. Just as one must not leap 
straight from the single generic form to the infinity of particulars, so on the 
contrary (???? ??? ?? ????????) when starting (as one must) from particulars 
one must not go straight to the one, but (in a clause admittedly difficult to 
translate and possibly corrupt; see Bury adloc.) first grasp the number of its 
species, and come to the one last of all. What follows is therefore not, as 
many call it, a third illustration of the process already outlined. It refers to 
the reverse process, and moreover seems to contradict the earlier instruction 
to start by identifying the generic form and only then go on to distinguish 
the species within it. The example tells how Theuth, reputed inventor of the 
alphabet (Phdr. 274 c-d), faced with the infinity of vocal sounds, first 
distinguished within them vowels, semi-vowels and consonants and their 
subdivisions, and finally reached the genus which he named * letters', uniting 
all the sounds into one. Bury adloc. says only * as in the analytical or deductive 
process, so likewise in the synthetic or inductive', but the simple antithesis 
between deduction and induction hardly fits. Stenzel gives the fullest 
explanation, but unless I have misunderstood Plato (which is not impossible), 
it does not quite correspond to his text. 
Let us assume men capable of articulate speech but not yet of writing. 
Theuth's first step was to find the largest group or species in the infinite 
number of phonemes (not necessarily of different kinds of phoneme, but 
including the daily repetition of what kinds there were: the apeira are 
individuals); his next, to divide these groups into their several components 
(e.g. vowels into a, e, i, o, u) until he * grasped the number of them and gave 
the name stoicheion [meaning both element and letter, p. 176 ?. ? above] to 
1 Analogous to the choice of different genera to produce different definitions of the Sophist. 
211 
15-2 
Philebus 
each and alP (i8c6). The letters, into which the groups are divided, are, one 
would think, infimae species, and it is to them that Theuth's method finally 
leads, not to the generic conception * utterance' (????) or its definition. To 
them he gave the generic name * elements', and this is evidently the 'one' to 
which one should come last. It conveys a clear conception, that speech is 
reducible to a limited number of atomic forms; but Theuth must have had 
this conception of stoicheion in his mind from the start. It was not the last 
arrived at. Though Plato does not seem to see it, there is no real reversal of 
the original procedure. At the very least one must say that Theuth assumed 
(the ?????? of i6d 2) a single generic notion to begin with, though only after 
he had seen that we cannot know a single set of letters on its own A8 c 7) did 
he discover (???????? i6di) the nature of stoicheion as a bond. (I make these 
comments with diffidence, but cf. Hackforth, PEP 26.) 
The fourfold analysis of everything B3C-26d). After satisfying Pro- 
tarchus that the good life must be * mixed', including both pleasure and 
knowledge, Socrates says that to justify his claim that intellect,  
knowledge and their kin deserve second prize he needs fresh resources. With 
a little self-mockery about his passion for * dividing according to kinds', 
he lays it down that everything in the world has a fourfold explanation. 
There are the unlimited and limit, already mentioned at 16 c, third the 
mixture of the two, and fourth the cause of their mixture. Unlimited 
includes qualities exhibiting an indefinite more or less, e.g. hotter- 
colder, stronger-weaker, lighter-heavier. Over against them are set the 
limiting factors of measure, proportion, ratio: double, for example, does 
not admit of more or less. All these involve number, and the imposition 
of the right numerical limit on an indefinite continuum (in Plato's terms 
their 'mixture', 2567) reconciles opposites and produces harmony. 
Examples are health,1 music (in the ranges of pitch and tempo), equable 
climate and so on. By slipping in the word 'right' (????, 2567) Plato 
virtually excludes from the category of limit the abstract, value- 
neutral concepts of pure mathematics, and confines it to the second of 
the two kinds of measurement described in the Politicus (pp. 169^ 
above), though obviously a medicine with three ingredients which 
1 P. has in mind the orthodox Greek medical theory, attributed to Alcmaeon, that health 
depends on a proper balance (????????* ??????) of the opposites hot and cold, dry and wet, 
bitter and sweet etc. in the body. See Symp. i86d and vol. i, 346. 
212 
Philebus 
depends for its effects on their being mixed in the proportions 1:3:7 
will not be improved if the proportions 1:4:16 are chosen because they 
represent a perfect geometrical progression.1 
Much has been written in attempts to find a place for the Forms in 
the fourfold classification. They have been detected in every category 
except the unlimited, and even in all four together. Since however the 
phrase 'all things which there now are in the all', which designates the 
subjects of the analysis, in my opinion clearly excludes the Forms, I 
shall not continue the chase.2 They might come under 'cause' (so 
Zeller), if the cause is to be thought of as external to the 'all',3 were not 
the cause explicitly said to be mind. (Nor again, since the mixture has a 
cause, can the Forms, being uncaused, play any part within it.) True, 
the Forms had a causal role in the Phaedo, but the most hardened 
unitarian must agree that Plato's thought has developed considerably 
since then. The development has been in the direction of reducing the 
elements of metaphor, about which Aristotle complained, and of the 
vagueness of a sentence like' Beauties became beautiful by the Beautiful, 
though whether by its presence or communion or how, I can't say' 
(Pho. iood). The Forms have become patterns only, formal causes 
which can create nothing without an agent at work to reproduce them 
in a medium. Soul, life and mind were dramatically introduced into 
Reality in the Sophist, in the Philebus mind is assigned its function as 
active cause, and the whole scheme becomes clear in the Timaeus, where 
the divine Craftsman, who is also Nous, creates the cosmos by  
reproducing, so far as the nature of the medium allows, the eternal, uncreated 
1 Aristotle saw the flaw in this. See p. 277 below. To P.'s Pythagorean friends limit as such 
was in the ' good' list (vol. 1, 245 f.), and words like ??????, ???????, ???????? regularly refer 
to due, or correct, measure. (See p. 169 above.) The association of measure and number with 
goodness was therefore natural to Plato, though if he had already drawn the distinction in PoL, 
to ignore it here was a pity. The distinction between practical and philosophical arithmetic is 
introduced much later, at 56 ?, in a different connexion. 
1 ????? ?? ??? ???? ?? ?? ????? B3 c) cannot refer to ?? ??? ???? ?? ???? ??????? ??????????? 
?????? E9c). Yet many have argued otherwise. For earlier views see Bury, lxiv-lxxiv and Ross, 
PTI132-8, and cf. Rodier, Etudes 79-93, wno found them in the mixed class, Friedlander, PL 
in, 324^ and Grube, P.'s Th. 301-4, who both identify them with limit. Dies (Bude ed. xciv) 
and Gentile (see Friedlander, PL in, 537 n. 37) saw them in all four. Hackforth (p. 34) and Rist 
{Philol. 1964, 227) agree that the classification excludes the Forms. 
3 This is at least doubtful. Cf. 64 c: 'What is it in the mixture which is at the same time most 
valuable and in the fullest sense the cause of such an arrangement commending itself to all of us?' 
At 23d the cause is a fourth class (?????) of the contents of the universe, and at 3oa-b it is 
?? ????? ???????? ????. See pp. 215 f. below. 
213 
Philebus 
Forms in the 'receptacle of becoming', variously described as a plastic 
material and as space.1 
Finally we have the application of all this to the case of pleasure. 
Its lack of limit (of which Philebus will boast), its encouragement of 
appetite to seek insatiably for more and more, is not a recommendation 
but a source of wickedness, and the imposition of limit, law and order, 
far from spoiling it, gives it a saving grace. 
The cause: cosmological and teleological arguments B6d~3i a). The  
mixture of the unlimited with limit is a 'being that has become' B7b, i.e. 
something in this world as opposed to the eternal Forms). It is accepted 
as axiomatic that nothing can be generated without a cause, and that 
'becoming' and 'being made' are the same. What, then, is the cause or 
maker of the mixture and the becoming? This question is highly  
relevant to the main aim of judging between pleasure and wisdom for 
second prize. The first goes to a life in the mixed category, pleasure falls 
within the unlimited, and a reader needs little perspicacity to see that 
Socrates will place wisdom (mind, thought, knowledge) in the  
remaining one. Since he has already emphasized the good effects of limit and 
its correlates measure, proportion, number, the competition is already 
virtually won, and from now on we are like readers of a detective story 
who have already guessed the outcome. The prime question, he  
continues, is whether the universe is at the mercy of chance, some random, 
irrational force, or guided by the regulating wisdom of Mind. Protarchus 
is shocked. The first alternative is sheer profanity, if not insanity. One 
has only to look at the cosmos, with its circling sun, moon and stars— 
responsible, as Socrates adds later, for years, seasons and months—to 
be convinced that it is the work of reason. This reply is a little  
unexpected from a man who has undertaken to defend hedonism. The view 
1 ????????? 50c, ???? 5238 (p. 265 below). The theme of Rist's article in Philol. 1964 is that 
in Phil, copies of the Forms, corresponding to the 'largeness in us' of Pho. i02d, and not the 
Forms themselves, enter the physical world and constitute the element of limit. (If, as appears 
likely, the Forms are already acquiring the numerical character which Aristotle ascribed to them, 
this would strengthen the identification.) The divine Mind is like a painter who paints many 
pictures of the same scene. The scene may be said to be in the pictures, but still remains unique 
and apart from them. R. claims that with this conception P. has solved the problems about Forms 
raised in Parm.y and he may be right, though I have queried some of his points on the Pho. in 
vol. iv, 354f. 
214 
Philebus 
that * Reason is King of heaven and earth', which Socrates complacently 
attributes to 'a consensus of all the wise', was not without powerful 
rivals among the natural philosophers, nor does Plato underrate them 
in book 10 of the Laws. Even the admired Theaetetus had his doubts 
{Soph. 265 c-d). By enthusiastically espousing it Protarchus has given 
away his case from the start, and one can imagine a satirical smile on the 
face of the listening Philebus. 
Microcosm mirrors macrocosm. Our bodies are composed of small 
portions of the same four elemental masses—fire, air, water and earth— 
as the universe, and draw on the body of the universe for their  
sustenance and growth. Likewise we have souls, and we cannot suppose 
that the Supreme Mind should have provided them for us yet failed to 
give soul as well as body to the macrocosm. Our souls indeed, like our 
bodies, are derived from their universal counterpart.1 After repeating 
that mind is the fourth kind, the cause, Socrates continues C0c): 'But 
mind and the power of thought cannot come into existence without 
soul, so you must say that in the nature of Zeus there appears a royal 
soul and mind by the power of the Cause.' This is not immediately 
clear, but Zeller and Hackforth have explained it well.2 Whichever came 
first, the Philebus and Timaeus expound the same cosmology and 
theology. Here there is an almost word-for-word coincidence,3 and we 
may make use of the fuller account at Tim. 29d~3ob. The supreme 
Cause is the divine Mind or intelligent Craftsman who created the 
cosmos. Knowing that nothing mindless would be as good as  
something intelligent, he gave it a mind. But nothing can have mind without 
soul, so he placed mind in soul and soul in a body. Soul {psyche) here is 
simply life, animation. The beasts have psyche without intelligence, and 
man cannot have intelligence without physical life. It is corporeal, 
visible nature, not the First Cause which is pure intellect, that cannot 
have mind without psyche.* Zeus in the Philebus passage is a literary 
1 This was a belief of Socrates himself according to Xen. Mem. ? .4.8. It can be traced back at 
least to Diogenes of Apollonia, who held that the soul is air and the air in us is * a small portion 
of the god* (vol. 11, 373f.). 
2 See, for both, Hackforth 56 n. 1. 
3 Phil. 30c ????? ??? ??? voOs ???? ????? ??? ?? ???? ?????????. Cf. Tim. 30b vous ?* ecu 
???!$ ????? ???????? ????????????? ??. That the elements of our bodies are ' borrowed * from 
the universe is also stated in Tim. D2?). 
4 Cf. p. 275 n. 1 below. 
215 
Philebus 
variation for the living and intelligent cosmos, borrowed from a 
pantheistic Orphic cosmogony.1 
Mind, then, may be said to be both immanent and transcendent, but 
the cosmic mind is only a self-projection of the other, who * being good 
and free from all envy, wanted everything to be as like himself as 
possible' {Tim. 29c). 
In all this Plato's debt to the Pythagoreans is obvious. Among  
notions which may fairly be ascribed to them, one may mention Limit and 
the Unlimited as primary principles and respectively good and bad, the 
importance assigned to number, ratio and harmony, the cosmos as a 
living and intelligent creature, and men's souls as fragments of the soul 
of the universe.2 
The psychology ojpleasure', pain and desire C1 d~36c). I avoid the  
heading * classification of Pleasures', though that may be said to start here, 
because, true to the conversational genre, Plato is very casual about his 
classifications. What he now calls pleasure covers only one kind, the 
* mixed'. Later come 'pure' pleasures, which even include the pleasures 
of learning. 
Pleasure occurs during the process of restoration or replenishment 
following on a disturbance of the bodily harmony causing pain. 
Examples are drinking when thirsty, cooling down when suffering from 
heat.3 Anyone experiencing neither deterioration nor restoration will 
feel neither pleasure nor pain. This state (which Socrates mentions here 
by the way, to be kept in mind for their future judgement of pleasure) 
would accompany the life of pure thought, which however is a life for 
the gods. To return, pleasure also accompanies the anticipation of  
something pleasant, as fear of something unpleasant may be called painful. 
This is a second kind, felt by the psyche independently of present bodily 
affections. To understand it calls for consideration of the nature of 
sensation, memory and desire, which are all involved. Some bodily 
1 See Guthrie, OGR 81. 
2 See the account of Pythagoreanism in vol. i, especially pp. 201, 207, 248, 289. 
3 In this account P. is unusually reticent about the pleasures of sex. ?????????? are mentioned 
at 65 c, and doubtless hinted at among the 'unseemly* pleasures of 46a-b with its tactful reference 
to Philebus, the 'shameful' of 66a and the 'greatest and intensest' of 63d. In the most general 
sense they satisfy a felt want, but would not be so appropriate to his description of physical 
pleasure as accompanying the 'refilling' of a deleterious 'emptiness' in the body. 
216 
Philebus 
changes do not reach the level of consciousness,1 but when soul and 
body are moved together in a common affection, this motion is called 
sensation. Memory is 'the maintenance of sensation',2 and recollection 
the process whereby a lost sensation is recaptured without further 
recourse to the body. 
The remarks on desire aim at proving, in a bafflingly roundabout and 
obscure way, that it is a purely psychical phenomenon. (' The argument 
denies that desire belongs to the body', 35 d.) Desire—thirst is the 
example—is for replenishment of what one lacks ('is being emptied of, 
34en), and is therefore for the opposite of what one is at the moment 
experiencing. Yet one must be aware of what one desires, and this can 
only be by remembering a previous experience of fulfilment, which, as 
we now know, is an activity of the psyche alone. 
There is an apparent contradiction here. To paraphrase 35a-d more 
fully, when the body is for the first time depleted, one cannot conceive 
of replenishment either by present sensation or through memory, since 
we have never experienced the pleasure of replenishment which the 
desire is for. Yet something in the thirsty (desiring) man must  
apprehend it C 5 b 6), and this must be his psyche, which can only do so by 
memory. How then, one must ask, does the psyche apprehend the 
object of its desire through memory on the occasion of the first 
emptying? Plato says nothing, but it can only do so if its memory is 
the pre-natal memory described in Meno, Phaedo and Phaedrus. This is 
not explained, nor is the difficulty seen by Protarchus, but I suggest 
it as a possibility because no alternative so far proposed seems  
satisfactory,3 and Plato does at least hint in this passage at the exalted 
1 In a repetition at 43 b, P. instances growth. Tht. 186 c also defines the objects of sensation as 
??? ??? ??? ??????? ???????? els ??? ????? ??????, and cf. Tim. 64a-c. 
2 Or preservation (???????). ?. has dealt with sensation and memory in Tht., and allows 
himself here a misleading brevity. He appears to be asserting the absurdity that the sensation 
itself persists, but cf. Tht. 164810: a man remembering what he has seen ???????? ??? ??? ???? 
?? ????. It is rather the image imprinted on the soul in the wax-tablet simile of Tht. 191 d. On the 
scope of ????????, which I have rendered ' sensation *, see pp. 74 f. above. 
3 Many of course believe either that the doctrine of anamnesis was never intended seriously 
(notably Moreau in IPQ 1959, 485), or that P. had by now abandoned it. Jowett (in, 533) 
thought its omission a sign of progress, * rendered all the more significant by his having occasion 
to speak of memory as the basis of desire*. Horn used the contradiction to support his attack on 
the dialogue's authenticity. Apelt thought that what is remembered is not the pleasurable process 
but the original bodily equilibrium (the state, one may add, in which neither pleasure nor pain is 
felt). But the use of ?????????, and the context, are strongly against this. As in Gorg. D96?), the 
217 
Philebus 
position of the soul as it has been described elsewhere. * The argument, 
by demonstrating that it is memory that leads us to the objects of  
desire, shows that every impulse and desire, and the source and principle 
(arckeI of every living creature, belong to soul' C5 d). 
Thus besides feeling pain when the body is depleted, and pleasure 
while its balance is being restored—processes closely connected with 
the destruction or preservation of life C5e)—there is an intermediate 
state2 in which we feel the pain of depletion but remember the pleasure 
which goes with replenishment. The memory gives pleasure if there is 
hope of future replenishment but doubles the pain if hope is absent. 
False pleasures C6c~44a).3 1. Pleasure may result from a hope, or 
belief, which turns out to be false.4 Socrates wishes to call this a false 
pleasure, but Protarchus objects. A belief may be true or false, but not a 
pleasure, because no one can think he feels pleasure without really 
feeling it. At first sight this seems irrefutable. * Pleasure cannot err since 
it is not cognitive.' 'The predicate is altogether inapplicable to the 
subject.' The pleasure that a man feels at the news that he has been 
appointed to an honourable and lucrative post is real, though if the 
news is false it will be short-lived.' The pleasure of hope is just as much 
pleasure when the hope is ill-founded, and due to false opinion, as when 
it is well-founded and based on right opinion.'5 This is exactly 
pleasure for which the desire is felt is the 'mixed* pleasure of drinking when thirsty. Hackforth 
(anticipated in this by Rodier, Etudes 98) simply says that desire does not occur at the first ???????, 
which appears to me to be ruled out by 35 a-b. (For Apelt and Hackforth, see the latter, p. 66 n. 1.) 
1 Jowett (in, 590) renders ???? by * moving principle*. Cf. Phdr. 245 c-d. 
* ?? ???? 35 ??. To be distinguished of course from the middle state of feeling neither pleasure 
nor pain because the body is in equilibrium C26-333). 
3 Gallop*s remarks on this section in PQ i960 are refreshingly independent and critical. 
4 The possibility of false belief was fully discussed in Tht. (pp. 106-13 above). Since P. is 
now after a different quarry, he accepts like anyone else the obvious fact that it occurs.  
Philosophers who earnestly discuss the existence of tables do not in fact doubt that they exist; they 
readily assume it when not philosophizing about it. For a modern discussion of the relation 
between pleasure and belief see the papers by Williams and Bedford in PAS vol. xxxin. 
5 A. Br?mond quoted by Friedlander, PL in, 539 n. 58; Grote, PL 11, 603f., Horn as cited by 
Bury, 206. Friedlander, Bury, Dies, Bude" ed. ciii-cv, Taylor PMW 421 f., and Rodier, Etudes 
113-28 are among P.*s defenders. More recently the question has aroused an extraordinary 
amount of interest. See Gosling in Phron. 1959 and 1961, Kenny in Phron. i960, Gallop in PQ 
i960, McLaughlin in PQ 1969, Dybikowski A) Phron. 1970, B) PQ 1970, Penner in Phron. 1970. 
One or two additional items are in the bibliographies of Dybikowski A), 147 n. 2 and Penner 
167 n. 3, and Gosling has resumed the discussion in his Phil., pp. 214-20. The criticism goes back 
to Theophrastus, who, says Olympiodorus, Opposes P. on the existence of true and false plea- 
218 
Philebus 
Protarchus's contention. As with beliefs, a man may be mistaken about 
the object which gives him pleasure, and the pleasure cannot then be 
called * right' or 'good' (??????), but what is false is only the belief on 
which it is founded, not the pleasure itself C7e~38a). 
One or two points from the text should be noted at once. 
First, at 36d and 38 a Protarchus denies the existence of false 
pleasures. At 4ob-c he admits it. So Socrates must have said something 
in between to make him change his mind. Second, Socrates himself 
asserts at 40 d that anyone who feels pleasure, however groundless, 
always really (ovtcos) feels that pleasure, though it may be based on no 
actual facts, past, present or future. Evidently the difference between 
them has been one of terminology: he and Protarchus meant different 
things by the word * false'. 
The analogy with belief is central. The beliefs * infect' the pleasures 
with their own condition D2 a). Beliefs have qualities, of which truth 
and falsehood are two. Pleasures can also be qualified: they have just 
been speaking of great or intense pleasures and their opposites. Now a 
belief, whether true or false, is none the less truly (ovtcos) a belief 
C7a 11). That is, a belief is judged true or false not in so far as it is or is 
not a belief, but with respect to the truth or falsity of its object. Why 
then should it be illegitimate to judge a pleasure by the same criterion? 
This is what Socrates proposes to do. By a false pleasure he does not 
mean an unreal one but one which arises from a false estimate of the 
situation, past, present or future. Belief and pleasure alike will be 
falsified in both senses when their falsehood becomes apparent, for they 
will cease to exist. Plato is using his own terminology, but has been at 
pains to make his meaning unambiguous. 
The account of pleasure in Republic 9 E83b-8yb) makes many of the 
same points as the Philebus, but here the Philebus shows an advance. At 
585 d-e Plato insists that the reality of a pleasure depends on the reality 
of its object: pleasure felt in the acquisition of what is more real (?? 
?????? ????, i.e. the objects of knowledge) is 'more really and truly' 
pleasure1 than pleasure felt in the less real. In the Philebus the reality 
sures. All are true, for if there is a false pleasure, he says, there will be a pleasure that is not a 
pleasure/ (Text in Rodier, o.c. 123.) 
1 ????? ?? xotl ??????????? ??????? . . . ????? ?????? E8561). 
2?9 
Philebus 
of the pleasure is clearly distinguished from the reality of its objects, 
while the words 'true* and * false' are retained in the specified sense, 
that is, with reference to the reality of the objects. 
Additional note. In discussing this type of false pleasures I have said nothing 
about the metaphor of the internal scribe and painter, or the idea that bad men 
are most likely to have their pleasures of hope falsified because the gods 
dislike them. Both have been discussed by others, and their introduction 
would only obscure what I hope has been a clear as well as brief indication 
of the main point, (i) The metaphor is not directly relevant to the problem 
of false pleasures, but to that of false beliefs. As such Runciman found it 
'more illuminating and more sophisticated than the unsuccessful analysis 
proffered in the Theaetetus\ But the question of justifying error is not raised 
here: its existence is admitted from the start C7b and e, 38b). See also 
Friedlander A11, 335 f.), who compares the treatment of false belief in 
Theaetetus, but adds that the possibility of error is a problem outside the 
scope of the Philebus, which need not be explicated there as it was in 
Theaetetus and Sophist, (ii) The immorality of the disappointed is in my 
opinion (not everyone's) a complete red herring. The argument requires 
only the fact of disappointed hopes, but we have seen more than once how 
characteristic it is of Plato's Socrates (and surely of the real one) to slip in, 
half humourously, a moral of his own. McLaughlin (in PQ 1969) says that 
* for Plato, falsity has definite moral implications'. In a somewhat specialized 
sense of * falsity' this is true (Rep. 382 a), though one cannot help recalling 
with some uneasiness the justification and extensive use of ?????? as an 
educative and political device in the Republic. (See vol. iv, 457f., 462.) 
2. At 41a Socrates suggests a second way1 in which pleasures (and 
pains) may be false. Again we are concerned with feelings of  
anticipation. The question is put: since pleasures (and pains) belong to the 
* Unlimited', i.e. admit of indefinite variation of magnitude, degree or 
intensity, is there any way of comparing them in these respects? It has 
been agreed that pleasures and pains are induced by a state of the body, 
that when its state occasions pain the soul desires the opposite pleasure 
which it is not experiencing, and the anticipatory pleasure exists 
simultaneously with the present pain. Now the distance in space at 
which an object is seen distorts our impression of its magnitude, and 
this is even more true of the temporal presence or distance of a pleasure 
1 He calls it the reverse of the first, presumably because in the first the pleasure was 'infected* 
by the belief, whereas in the second a false belief is the result of an illusory feeling of pleasure. 
220 
Philebus 
or pain. The amount by which the anticipated pleasures exceed or fall 
short of the actual (??? ?????, 42 b 8) must be reckoned unreal. 
In this highly artificial argument Plato is not expressing his full 
thoughts. The distortion due to distance is made to sound inevitable, 
and there is no mention of counteracting it by an art of measurement. 
Contrast the hedonistic argument in the Protagoras C57a-b), where 
also the analogy with spatial distance has been invoked: * Since our 
salvation in life lies in the correct choice of pleasure and pain—more or 
less, greater or smaller, nearer or more distant—is it not in the first 
place a question of measurement, consisting in a consideration of 
relative excess, defect or equality?' So to live a successful life on 
hedonistic principles demands the acquisition of a techne, that is, of 
knowledge, which transforms it into something very different from 
hedonism as vulgarly understood. Later in the present dialogue the 
importance of measure and techne will loom large, and be the means of 
putting pleasure in its proper subordinate place. 
3. The third type of false pleasures mentioned does involve the 
'impossibility* that one can suppose one is feeling pleasure when one is 
not. * False* here means unreal. The inconsistency is probably due to 
Plato's wish to account for all current theories about pleasure which 
he thought wrong. One of these was certainly the belief that the 
pleasantest (???????) life was one of calm and tranquillity undisturbed 
by either pleasures (in the commonly accepted sense) or pains, the state 
which Democritus called euthymie and Epicurus would soon make 
famous as ataraxia. It may have been more widely known as a bit of 
proverbial wisdom.1 Socrates adapts it to his physiological explanation 
of pleasure by calling it the state in which the body is undergoing 
neither of the processes of deterioration and renewal, or at least (if as 
the Heracliteans say such stability is unattainable) not to a perceptible 
1 For Democritus see vol. n, 492 k Antisthenes has also been suggested, for whose  
condemnation of pleasure see vol. m, 307. But one recalls also the reported contentment of the aged 
Sophocles at his release from the tyranny of sexual desire (Plato, Rep, 329 c), and especially the 
sentiments of Amphitryon in Euripides, H.F. 503-5: 
????? ??? ?? ??? ????, 
?????? ?* ???$ ?????? ??????????? 
?? ?????? ??$ ????? ?? ??????????. 
Cf. Phil. 43d ^s ??????? ?????? ????? ?????? ????????? ??? ???? ?????? and 44b ??? ?'  
???????? ? ??? ????? ... ??? ???????????????. 
221 
Philebus 
degree. People do say and believe that in such a state they are  
feeling pleasure, but since positive pleasure and absence of pain are two 
different things, their belief that they are feeling pleasure must be 
false D4a9~io).1 
Socrates's constant reference to pleasure as such as if it consisted 
solely of physical pleasures and their anticipation is an irritating feature 
of the Philebus, scarcely redeemed by the partial justification suggested 
on behalf of the anti-hedonists at 446-45 a (that the nature of anything 
appears most plainly in its extreme or intensest forms and bodily 
pleasures are the commonest and greatest) or by the fact that Socrates 
will shortly show that he himself does not so limit it.2 One may 
guess that the explanation is again historical. Current controversy 
between hedonists and anti-hedonists did concern indulgence in 
physical pleasures, and the conception of these as disturbances lay 
behind such dicta as the 'give me madness rather than pleasure' of 
Antisthenes.3 
Are there any true pleasures? Description of pleasure-pain compounds 
D4b-~5oe). Considerations like the foregoing have led certain  
philosophers4 to deny the existence of pleasures altogether. There are not, 
they say, three states, painful, neutral and pleasant, but only two, pain 
and relief from pain. Though this is not quite true, they show a severely 
puritanical zeal* which is commendable in its detestation of the un- 
1 A repetition of the argument of Rep. 5830-85^ 
2 It is fascinating to compare Ryle's essay on pleasure in Dilemmas (pp. 54-67) with some 
ancient views. For instance he agrees with Aristotle against P. that pleasure is not a process (p. 60). 
Of pains he too speaks apparently universally as ' the effects of such things as the pressure of a 
shoe on a toe', and the sense of' pains' at the top of p. 67 appears to be the same. Yet he would 
not have denied us the right to speak of the pain of bereavement or parting, and the Greek ????? 
had similar scope. 
3 Fr. 108 Caizzi. Notice the connexion between profligate pleasures and madness at 45 c 
4 Described as * recognized experts in the study of nature' D4 b 9). Opinions as to their identity 
have differed widely. (Bignone's argument for Antiphon in Studi 221-6 is one that seems to have 
been forgotten.) Perhaps the non liquet of Wilamowitz {PL 11, 272), Jowett (in, 542) and Hack- 
forth (p. 87) must be accepted, but more recently M. Schofield has vigorously revived the case for 
Speusippus, whom some earlier critics had rejected (Mus. Helv. 1971, 2-20 and 181), though 
Taylor too thought the description 'exactly fits Speusippus' (Timaeus, 456). K. Bringmann, on 
the other hand, in Hermes 1972, has argued equally vigorously for Heraclides Ponticus. 
5 Not, I hope, too wide of the mark as a translation of ????????? ?????? ??? ??????* in this 
context. I now see that 'Puritan spirit' occurred to Bury too. (But Schofield suggests a reason 
why ????????? might be peculiarly appropriate to Speusippus.) 
222 
Philebus 
healthy power and 'witchery' of pleasure. They concentrate on bodily 
pleasures as the most obvious and greatest. Of these in turn the greatest 
are those preceded by the greatest desires, and are therefore felt  
predominantly by the ailing in body or mind, in the satisfaction of the 
morbid desires of the sick or those who lack all self-control. Starting 
with the relief of an itch by rubbing, Socrates calls all such pleasures 
' unseemly', and without specifying its causes vividly describes the mad 
delight of the fool or profligate, 'dying with pleasure' and calling  
himself the happiest of men. 
Such pleasures are inevitably mixed with pain, and originate in the 
body alone. A second type of mixed pleasures, Socrates recalls, has been 
discussed already, namely those where a single pleasure-pain complex 
is formed from the pain felt in a present bodily depletion combined with 
the pleasure of the mind {psyche) in anticipating future replenishment. 
Thirdly, there are mixtures of pleasure and pain in which the psyche 
alone is concerned. These are the emotions—anger, fear, longing, grief, 
love, emulation, malice D7 e). They are pains, yet also fraught with 
immense pleasures, as Homer speaks of wrath as 'sweeter than honey in 
the human breast'. Even grief and longing, Socrates claims, are mingled 
with pleasure, and he instances the effect of hearing a tragedy, or, in the 
case of malice at another's ridiculous misfortunes, a comedy, with no 
hint of any difference between personal and vicarious experience of 
tragic or comic situations. He goes on to analyse the psychology of the 
pleasure-pain combination arising from such causes as a malicious 
sense of the ridiculous or the Schadenfreude which we cannot help 
feeling when even our friends make fools of themselves. Then realizing 
that he has got into a digression, he sets aside an examination of the 
other emotions until they have settled the main question raised by 
Philebus's attitude to pleasure. 
Punles. At places in the Philebus the threads get bewilderingly  
entangled, and Plato's application of the same descriptions to ideas which 
sometimes are, but sometimes are not, the same is certainly no help in 
unravelling them. Nor can one resist a sneaking doubt whether in every 
case the attempt is worth while. His language seems almost intentionally 
mystifying, and moreover what he is really driving at only becomes 
223 
Philebus 
fully clear in the concluding part of the dialogue. As far as I can see, in 
the most recent sections there are two outstanding difficulties. 
i. What is the relation between various states described as neutral, or 
as 'intermediate lives'? 
(a) 32 c Anyone undergoing neither deterioration nor restoration of 
body feels neither pleasure nor pain. A life of uninterrupted thought 
would be of this kind, and would be fit for gods. 
(?) 35 b—36b. When one experiences a present pain of depletion but 
entertains a pleasurable hope of repletion, one may be said to be pained 
and pleased at the same time. 
This is obviously different from the neutral state of feeling neither 
pleasure nor pain, but like (c) is called an intermediate sort of life, one 
'in the middle'.1 
(c) 42c If no perceptible depletion or replenishment is taking place 
in the body, people feel neither pleasure nor pain, but think they are 
feeling pleasure. This is a case of false pleasure, and is called the  
intermediate, or middle, life.2 
Setting aside (?), there is nothing in the descriptions to suggest a 
difference between (a) and (c). Yet (a) is exemplified by a life of pure 
thought unmixed with pleasure, which, it now appears, was only 
rejected as the best for man because it is beyond his reach, whereas the 
false pleasure of (c) sounds more like the sensation of a once-hungry 
man of sound digestion peacefully relaxing after a good meal, (c) refers 
to the common man, (a) to a super-philosopher who has succeeded in 
the philosopher's aim of'assimilation to God' {Thu 176b). 
2. How do the 'puritans' deny the existence of pleasure and at the 
same time detest its evil influence? 
Their introduction at 44b-d is made to sound like a straight follow- 
on from what has preceded. These 'enemies of Philebus' assert that 
what he and his like call pleasures are simply escapes from pain. One 
expects this to refer to case i(c), where neither pain nor positive  
pleasure is being felt because the body is in a state of equilibrium. Yet they 
go on to say confusingly that the very attractiveness of pleasure is not 
1 ???? ????? 35 dc>, ?? ???? ey. 
2 ? ????? ????, 43e^· At 55a the life without pleasure or pain becomes once again the life of 
pure thought as in (a). 
224 
Philebus 
pleasure (sic, 44C-d) but illusion, and take as their examples not the 
placidity resulting from this physical balance, nor the quietist ideal of a 
life without pain, but the most violent and profligate of pleasures, which 
certainly fit Socrates's initial description of pleasure as accompanying 
the process of satisfying what is felt as a bodily need. Socrates goes on 
to describe these, not as states neither pleasurable nor painful but on the 
contrary as mixtures of pleasure and pain.1 
The upshot of it all is perhaps no more than this. 'Pleasure' stands 
for what worldlings like Philebus call pleasure, that is, the physical 
pleasures only, whose character has been more fully expressed through 
the mouth of Callicles in the Gorgias.2 Our severe moralists refuse it the 
title of pleasure and condemn its false allure: its devotees are not  
enjoying pleasure but endlessly seeking relief from the tyranny of their 
pathologically inordinate desires. The unmarked transition from ' relief 
from pain' as a calmly neutral state to 'relief from pain' as indulgence in 
the most intense and exciting of pleasures seems monstrously illogical 
today. It is probably to be explained, though hardly excused, by the 
familiarity of Plato and his original readers with current ethical 
theories and controversy. He has already said in the Republic E84 c) 
that 'most—and those the most intense—of the so-called pleasures that 
reach the psyche through the body, are of this kind, relief from pains'. 
True pleasures Eoe~53c). Socrates does not conceal his sympathy with 
these opponents of vulgar hedonism. Their only mistake was to confine 
the whole concept of pleasure to what is in fact a travesty of it. Pure 
pleasures, unmixed with the body's pains and ailments, do exist and 
truly deserve the name of pleasures.3 They include the enjoyment of 
1 One is tempted to throw back at S. the question he himself asked in Rep, E83c): 'Is it 
possible that what is neither of two things should be both?' It is difficult to decide at what point 
he ceases to speak for the puritans and reverts to his own views, but they are at least credited with 
the advice to seek the true nature of anything not in a mean but in its extreme forms, e.g. if you 
want to know what hardness is, take the hardest thing you can find D4d-e). 
* It is the life of the charadrios-bird, continually eating and excreting. (At Tim. 726-73 a the 
coiling of the intestines is explained teleologically as a device to counteract the natural human 
propensity to overeat. See pp. 313 f. below.) In connexion with Phil. 46 a one may note the 
transition from scratching an itch to sexual activity. See Gorg. 494 b-c. 
3 Throughout this passage S. uses 'pure' or 'unmixed' (???????, ????????) and 'true', 
'genuine' or 'real' (???????) as synonyms. He argues briefly for this at 52d~53c. Taylor's 'true 
to type' comes close to what P. has in mind (PMW qrj). 
16 
225 
GHG 
Philebus 
beautiful colours, shapes, and sounds, and even, though on a lower 
level, scents.1 The shapes must not be those of representational art, but 
abstract geometrical forms, either plane or solid, and the sounds a 
series of single pure notes, for there must be no element of association 
or comparison in a pure pleasure. Such pleasures approach the divine. 
It is the apotheosis (or would an art critic say the reductio ad absurdum?) 
of the peculiarly Greek preference, exemplified par excellence in the 
Pythagoreans, for 'the intelligible, determinate, measurable',2 the keen 
appreciation of mathematical form, symmetry and proportion which 
pervades classical architecture, sculpture, pottery and literature, and 
which made such a tremendous appeal to Plato that he prescribed a long 
and arduous course in mathematics as an essential part of a statesman's 
training. (Politics, one must admit, was the field in which the Greek 
instinct for due measure and avoidance of excess was least conspicuous.K 
These contemplative pleasures satisfy two conditions: they are 
untainted by the pain of a previous perceptible lack, and the beauty of 
their objects is intrinsic. The description of these as beautiful 'not 
relatively, in comparison with others, but in and by themselves' 
E1 c, d) is interesting, and led Hackforth to suppose them to be' perfect 
particulars of the Idea of Beauty, its fully adequate expression to sense' 
{PEP 99). I do not see that his second designation coincides with the 
first, nor did Plato ever think of visible surfaces and solids, 'produced 
by the carpenter's lathe, rule and square' E1c), as perfect. At 62b he 
actually calls these 'human' circles and straight-edges 'false'. The 
language, it is true, resembles that reserved elsewhere for Forms, and 
may be another indication that in these later dialogues Plato shows more 
respect for the sensible world than previously. Hackforth reminds us of 
what the Phaedrus says about visible beauty, but that contains no 
suggestion that earthly objects can be anything but imperfect copies of 
1 'Most scents', says S. E1b4). Hackforth suggested that those excluded were only enjoyed 
by contrast with preceding unpleasant odours, Taylor that 'of course' he was excluding smells 
of food or the scent of the female perceived by the male {PEP 98 n. 1, PMW 426 n. 1). At Rep. 
584 b odours are mentioned without qualification as an outstanding example of pure pleasures, 
and I doubt if ???????* here has any special significance. Unpleasant smells naturally do not give 
pleasure. 
* The words are from E. Fraenkel's inaugural lecture Rome and Greek Culture 25. 
3 Aristotle is a true follower of P. when he castigates those who deny that mathematics has 
anything to say about goodness and beauty.' The chief forms of beauty are order, symmetry and 
limit, and these the mathematical sciences demonstrate to an especial degree' (Metap/i. 1078331). 
226 
Philebus 
the Form of Beauty. In any case the present concession to the sensible 
world is minimal. We are not to admire a beautiful landscape or girl 
(who may be beautiful compared to a monkey but not to a goddess, 
H.Maj. 289 a-b), but solely mathematical drawings and models, or 
series of'single pure notes', whose pitch the Pythagoreans had shown 
to be essentially a matter of number. And mathematics is the study by 
which the philosopher 'rises out of this transient world to a grasp of 
reality' {Rep. 525b). 
Pure pleasures include also those of learning E1 e). The lack of 
knowledge, which learning fills, or the loss of it through forgetfulness, 
are not perceived as painful (though pain may be caused incidentally by 
the practical consequences of forgetfulness, 52b 1). These pleasures are 
only for the few. 
Having distinguished pure from impure1 pleasures, Socrates,  
reverting to his earlier diairesis of all things, assigns the impure, which  
(following the diagnosis of the puritans) he now also calls vehement, to the 
Unlimited, subject to the measureless range of more and less, and the 
pure to the Limited or measurable class. Measurable pleasures are 
evidently to be regarded as stimulated by measurable objects.2 
Finally, anything in its purest, least adulterated form is more truly 
that thing than even a larger quantity of it contaminated with  
something else. So the pure pleasures felt in the beauty of geometrical shapes 
and in acquiring knowledge are more truly pleasures than those,  
however intense, that are mixed with pain.3 This is illustrated by an 
analogy: Even a small patch of pure white colour is more truly white 
than a large surface on which white is mixed with some other colour.4 
Given Plato's presuppositions this seems fair enough, and one cannot 
read the Philebus at all unless one accepts that it is an exposition of his 
credo rather than a defence of it. A question that one can legitimately 
1 Instead of simply 'mixed', at 52c Plato uses for the first time the adjective ????????* with 
its rhetorical overtones of physical, moral or ritual uncleanliness or miasma (Pho. 80 b). 
2 Though how scents, as well as spheres and cubes, can be measured is not explained. (In fact 
they cannot. See Tim. 66d-e.) 
3 There is no change here from the Republic, where any pleasure except the wise man's is 
???? ????????? ???? ??????. 
4 In the case of colour this seems to be true, and I do not see the relevance of Gallop's  
introduction of parti-coloured objects in PQ i960, 340 n. 6. Crombie's complaint (EPD 1, 260) is that it 
is not a valid analogy to pleasure. 
227 
16-2 
Philebus 
ask, because it falls within the framework of Platonism, is: if the pleasure 
of learning is not preceded by any sense of lack, what has become of the 
philosopher's eros, his passionate longing for truth? Jowett (in, 53) 
speaks of its absence as marking a different stage of Plato's thought 
from the Symposium and Phaedrus, but it is not absent: love of, or 
longing for, truth (???? ??? ???????) is mentioned as a natural faculty 
of the psyche at 58c!. 
Socrates is now beginning to unmask his batteries, and we can  
foresee the inevitable outcome of the fight. In introducing the topic of the 
place of pleasure in the good life, he started from the conception of it 
held by a hedonist like Philebus, which enabled him to place 'pleasure' 
in the Unlimited class. Now he has made it clear that these are not the 
only, nor even the genuine and desirable pleasures (though he will 
shortly switch back, without warning, to the earlier use of the term!). 
True pleasures belong to Limit, they observe due measure and are 
better (??????? 53c2) than the others, and chief among them are the 
pleasures of contemplating geometrical forms and acquiring  
knowledge. The original admission that pleasure as well as intellectual 
activity must play a part in the good and happy life takes on a rather 
dubious air when the pleasures that contribute to it turn out to be the 
pleasures of intellectual activity. 
Pleasure as process and means E3C-55C). This is an unsatisfactory little 
argument, soon to be refuted by Aristotle. That pleasure is a process 
(genesis, a coming or bringing into being as opposed to a completed 
state of being) is not argued but laid down as a premise on the  
authority of certain pundits, 'to whom', says Socrates, 'we should be  
grateful'.1 No doubt for Plato it seemed a necessary consequence of its 
association with processes of bodily restoration and its cessation when 
restoration was complete. A generative process, he goes on, like the 
instruments it uses, is never an end in itself, nor good for its own sake: 
the goodness is in the product at which it aims. So pleasure, being a 
process, can never be good in itself, and Socrates's informants laugh at 
1 Various guesses at their identity have been made (e.g. Speusippus, Hackforth, PEP 106), 
but as S. goes on they come to sound suspiciously like his troublesome relative in H.Maj. or 
Diotima in Symp., i.e. nothing but a vehicle for his (or P.'s) own views. See vol. iv, 176 and 385. 
228 
Philebus 
its devotees, who are happy to put up with hunger and thirst in order 
to enjoy what they regard as the pleasure of eating and drinking, a life 
of continual demolition and rebuilding, rather than the life of the purest 
possible thought containing neither pleasure nor pain. 
Plato's thought is less confused than his exasperating language. 
'Pleasure' has jumped back to mean the Phileban pleasures which we 
have now been taught to regard as no true pleasures, while the Socratic 
pleasures of the mind, just represented as the purest and best, revert to 
being no pleasures at all. But with this sorted out, it does not follow 
that because the satisfaction of hunger is a process, occupying time, and 
approaching its fulfilment by stages, the pleasure accompanying it is 
also a process (presumably progressing inversely, since it ceases at the 
end, as if one were bound to enjoy the hors d'oeuvre more than the 
sweet). Rather, as Aristotle was quick to point out, pleasure is not any 
sort of motion or change 'because no change is complete at any and 
every moment... whereas the form of pleasure is complete at any time'. 
' It is a whole. You could never find a pleasure such that if it lasted 
longer its form would be brought to completion.'1 
Analysis of knowledge E5C-59C). After pleasure, it is only right that 
reason and knowledge should be scrutinized, to discover whether some 
kinds are truer or purer than others. For this purpose 'knowledge', like 
'pleasure', is given its widest application, even including guesswork and 
doxa, which in earlier dialogues have been strongly contrasted with 
knowledge. No change of doctrine is involved, for considered as 
knowledge they are soon shown to be neither true nor pure; but as with 
pleasure, for dialectical purposes Plato needed a single general term to 
be the subject of division. The whole field of knowledge is first 
divided into practical, or technical, and educational or cultural. Of the 
former, some branches (e.g. building) make great use of measuring aids 
and instruments, whereas music,2 medicine and others rely on trial and 
1 See Arist. EN 1174313^14, 11521312-15, 1153312-15. The allusions to P.'s Phil, are 
unmistakable. 
2 It may seem surprising to find music in this category, in view of the Pythagorean success in 
demonstrating the mathematical basis of music, and P.'s own teaching about harmonia involving 
measure and its exploitation in Tim. But apart from the fact that he was not entirely uncritical of 
Pythagorean methods {Rep. 531 b-c), he is here speaking of execution, not theory; and 56a 
shows how much room there still was for a hit-or-miss technique in contemporary musical 
229 
Philebus 
error and guesswork. So the arts themselves subdivide into the more 
and less exact. Most exact of all is arithmetic, with its kindred, measuring 
and weighing. But must we not divide this art of number in its turn, 
into practical and philosophical? The man who uses number in his 
practical work must take as his units physical things like cows or 
bricks which are never precisely equal (nor of course are they in 
philosophic eyes purely unitary), whereas the pure (philosophic) 
mathematician deals solely in abstract units all absolutely equal and 
identical.1 
Once again (cf. p. 167 above) we see that the dialectical framework 
is no unrealistically rigid one. It would be difficult to make a  
straightforward tabulation of these divisions on the lines of the Sophist (pp. 124 f.). 
Perhaps Plato had come to see the inevitable disadvantages of such a 
stringent scheme, 'easy to indicate but hard to practise' A6b). Here 
arithmetic turns up on the practical side of the dichotomy, but one of its 
subdivisions clearly belongs to the theoretical, and Socrates adds for 
good measure that there are plenty more such 'twin pairs' among the 
arts, though united under a single name.2 Yet regarded not as a 
diairesis but as a single scale of ascending degrees of precision the 
passage is perfectly in order: music and its kin, building, practical 
calculation, pure arithmetic or theory of numbers, dialectic. 
It only remains to equate the precision or exactness of a science with 
the degree to which it is 'purely and truly' a form of knowledge, and 
this is soon done. If after attending a meeting I said ? know there were 
at least 100 people there', Plato would say that my knowledge was not 
only less precise, but less properly to be termed knowledge, than if I 
knew there were 106 because all present voted and the voting was 60 to 
46. It might therefore appear that mathematics or any science carried 
on in its philosophic mode is the most exact form of knowledge. But this, 
performance. Cf. Rep. 531 a-b. On medicine Plato is in line with some contemporary medical 
opinion. The author of On Ancient Medicine writes (ch. 9, i. 588 L.): * A certain measure must be 
the aim, but you will not find any weight, number or other standard, reference to which will make 
knowledge precise; there is nothing but the body's sensations.* 
1 On this see vol. iv, 523 n. 1, where a reference to the ??????* ?????? of 56dio might have 
been apposite. 
* This is further evidence, in conjunction with the Crat., that the statement at Rep. 596 a, about 
positing a single form for every set of things to which we give a single name, was not intended 
to be taken au pied de la lettre. For this see pp. 25 f. above and vol. iv, 550. 
23O 
Philebus 
says Socrates, would be to neglect dialectic, the truest and most exact 
of all. It is the old story. No science can be exact which deals with ever- 
changing material, and only dialectic has for its objects the absolutely 
real, unchanging and separate. Only in them, and what is most akin to 
them,1 can we find stability, purity and truth: in other words, the 
transcendent Forms which we have known since the Phaedo.2 
If the conception of'twin arts', and their relation to dialectic, is at all 
obscure, one has only to turn back to some pages of the Republic, of 
which this passage is no more than a brief summary. (Plato lived up to 
the proverb he is just about to quote, that what is worth saying is worth 
saying twice or thrice.) There in book 7, 523a~32b, we have a full 
description of the two sorts of arithmetic, geometry, astronomy and 
harmonics, one directed to practical use—in commerce, war,  
navigation, agriculture and so on—and the other leading the mind upwards to 
perfect and changeless entities like numbers, regular figures, the 
mathematical relations embodied in the notes of a scale, and the ideal 
motions of which the revolutions of the heavenly bodies are the  
physical counterpart. There, as here, all are propaedeutic to dialectic, the 
study whose final aim is nothing less than a grasp of the Form of the 
Good through pure thought E32a-b).3 
A short digression on rhetoric sparked off by Protarchus, who like 
many young men has been impressed by Gorgias's claim that it is the 
finest of all arts, is the occasion for Plato to repeat his criticism of it  
elsewhere as relying on probability instead of truth, and to give a brief 
reminder of the concept of doxa. Not only rhetoricians, he says, but 
even natural philosophers rely on this, for of the changing world of 
sense there can be only opinion, not knowledge. Even when, in the 
Timaeus, he himself thought it worth while to devote a long and 
1 Hackforth thought (PEP 122 n. 2) the * closest kin' were the objects of astronomy. But P. 
is talking not of astronomy, even in its philosophical form, but of the objects of dialectic. If the 
insertion is significant at all, it probably refers to souls, or rather minds, described as akin to the 
Forms in Pho. and admitted to the realm of the completely real in Soph. 
a ?? ??? ???? ?? ???? ??????$ ?????? E9c4) ls tne familiar formula. (Cf. 58a, 6id-e and 
p. 141 n. 2 above.) With ???????, ?????????$, ???????? cf. the same words used of ???? ?? 
????? at Symp. 211 e. See also additional note on p. 232. 
3 Though the objects of the other philosophical sciences are also fully real and changeless 
their practitioners differ from the dialectician in that they do not use pure thought alone, but are 
still dependent on objects of sense (visible figures, stars and the like) to set them on their way. 
For the contrast see Rep. 510C-11 c, and cf. vol. iv, 509 f. 
231 
Philebus 
detailed disquisition to the natural world and its creatures, he never lost 
sight either of its subordinate place on the ontological scale or of the 
goodness which it nevertheless exhibits through being modelled on the 
Forms by the divine Mind—another point emphasized in the present 
dialogue. 
ADDITIONAL NOTE ON BEING AND BECOMING 
IN THE 'PHILEBUS* 
At 16C9 I have translated ??? ??? ????????? ????? 'all things that are ever 
said to be* (p. 317),and taken it to refer to the phenomenal world. Plato uses 
????? loosely of ????????? when he is not emphasizing the distinction. As I 
remarked in connexion with Tim. 52a (vol. iv, 495), it is a difficult word 
to get rid of. In Phil, we have also the phrases ??????? ets ?????? B6d) for 
the coming-to-be of a thing in this world from the blending of limit and 
unlimited, and ?????????? ????? at 27 b; and at 64b he says that without an 
admixture of reality nothing 'can truly become nor be a thing that has 
become' (??? ?? ???????? ???' ?? ????????? ???). Hackforth is right in 
saying {PEP 49 n. 2) that we should not read too much into such  
expressions. Of the other examples which he quotes, Tim. 35 a (?????? . . . 
??????????) is particularly apposite (and cf. 31b 3 ??????? ????? ??? ??' 
?????), for in no other dialogue is the contrast between Being and Becoming 
so uncompromisingly drawn B7d-28a). One cannot agree with those who 
see the expressions as marking a radical change in Plato's philosophy, putting 
sensible things on the same ontological level as the Forms. At the same time 
Bury too was right to remark, on p. 211 of his edition, that this apparent 
weakening of the barrier between Being and Becoming occurs especially in 
the later dialogues (note the sources of Hackforth's parallels), where it seems 
to be a question less of contrasting Being with Becoming than of  
distinguishing grades of Being. This accords with the shift of interest towards the 
temporal world which became noticeable in Pol. and reaches its climax in 
Tim. I suspect however that the linguistic variations were largely unconscious. 
But see also Solmsen's judicious note in Aristotle s System, 39 n. 79, and 
p. 233 n. 3 below. 
Composition of the mixed life: pleasure loses second pri^e E9d-66a). The 
mixed life of knowledge and pleasure is, as they agreed at the beginning, 
the best, rather than a life of either alone,1 as Socrates and Philebus had 
1 Enjoyment of pleasure with no trace of mental activity, neither remembering past nor looking 
forward to future pleasure, nor even able to reflect that one is enjoying it B1 b-c), sounds like a 
232 
Philebus 
earlier been contending. Its ingredients being now prepared, it remains 
to decide how to blend them, and award second place to one or the 
other. Each has been found to admit of several varieties, some more 
genuine and true to type, others adulterated. The question as Socrates 
sees it is whether all kinds of each should be admitted to the good life, 
and if not, which. Take knowledge first. The thought of a man knowing 
the true nature of Justice and the other Forms, but none of the practical 
arts, makes Protarchus laugh. Life itself demands that the doors be 
flung open and all forms of knowledge allowed to pour in, the inferior 
with the pure. Is it the same with pleasures? The answer, given as a 
reply by Pleasures on the one hand and Thought and Intelligence on 
the other to the question whether they would like to live together, is 
that the Pleasures would welcome all knowledge, both in general and of 
themselves,1 whereas Thought would recoil from any pleasures which 
were not pure and true, or at least attendant on health, temperance and 
all virtue. The vehement and intense pleasures that go with folly and 
vice simply harry and destroy it with their frenzy.2 
The 'necessary' pleasures which Plato (somewhat grudgingly, one 
feels) admits F2 e) will be those that go with nourishment and  
procreation. These cannot be ensured without pleasure, and the life being 
prescribed for is that of a 'living body9 F4by), but he has not  
concealed his view that if we could rid ourselves of the body and live by 
thought alone we should be as gods. The emphasis may have changed, 
but the doctrine is still substantially that of the Phaedo. 
As a third and final ingredient Socrates adds reality to the blend. To 
treat it as a component sounds odd, but his next sentence makes the 
meaning clear: 'Anything in which we are not going to mix reality will 
not really come into being nor be a thing that has come to be' F4b).3 
purely logical construction. Yet Aristippus seems to have come near to recommending it as a 
practical aim. He is said to have defined the goal of life as momentary experience of pleasure with 
no thought of past or future. Memories and hopes were alike irrelevant, for the past no longer 
existed and the future might never be. Only the present moment was ours. (Frr. 207 and 208 
Mannebach.) " This is probably aimed at Aristippus. See previous note. 
2 A repetition of the teaching of the Phaedo, Cf. especially what is said about bodily desires 
at 64 d and 65b-d. Again at Tim. 86b-c excessive pleasures and pains are called the greatest 
diseases of the psyche, under whose influence a man cannot see nor hear aright and is incapable of 
reasoning. 
3 To translate ????????? and its inflections without clumsiness is always a problem, for there 
is no corresponding single English word. Jowett has 'nothing can truly be created or subsist', 
233 
Philebus 
This is probably intended as a reminder of Plato's increasing concern, 
manifested also in the Politicus, that his principles must be capable of 
being put into practice. 
The good then is in the mixed life. That is its dwelling or lair F1 a-b, 
64 c), but we must still ask what is the most valuable element in the 
mixture, and makes it universally desirable, and after that, whether it is 
more akin to pleasure or intelligence in the whole scheme of things 
(?? ?? ?????, 64C9). But this is easily answered. Everyone knows that 
the goodness in any mixture depends on correct measure or proportion. 
This at once unites goodness with beauty, for that too is a matter of 
measure and proportion like any kind of excellence (arete).1 Reality2 too 
is in the mixture, and without further ado is declared to be part of that 
which makes it good; so the good is a trinity in unity, combining 
beauty, proportion and reality.3 All these are present in intelligence and 
thought, whereas pleasure is a cheat (in love, reputed the greatest 
pleasure, even perjury is forgiven) and unmeasured in its intensity, and 
the greatest pleasures are either ugly or ridiculous, banished for shame 
from daylight to the hours of darkness. Clearly pleasure cannot be 
given second place as more responsible than intelligence for the  
goodness in the good life. 
In this brief argument, if such it can be called, the reader is rushed 
breathlessly along to the ultimate Q.E.D. of the whole dialogue. Why 
does it follow that because reality is in the mixture it should be part of 
that which makes it good? What about those pleasures which are no 
less in it? Protarchus's by now enthusiastic cooperation makes it clear 
that in deciding the issue between pleasure and thought only the 'false', 
a better rendering than Hackforth's. P. has in mind of course his standard contrast between 
????????? and ????, impermanent physical things and the eternal Forms. The former are not 
?? ????, they have a share in reality through participating in the Forms. Without any reality at 
all, a man (say) would never be born, nor grow to be that transitory thing (?????????) which we 
call a man. Phrases like ????????? ???, ?????? ??$ ?????? B6 d) and ?????????? ????? B7b) 
certainly do not indicate that P. is now promoting ????????? to the full status of ????. The 
difference between them is emphasized throughout the dialogue. (See additional note on p. 232.) 
1 This, like so much else, is repeated in 77m. (87 c): ??? ?? ?? ?????? ????? ?? ?? ????? ??? 
???????. 
* ??????? again, usually translated here F469) as * truth \ I have thought it better to make it  
uniform with the occurrence at b 2 to which P. is referring, where the meaning * reality' is uppermost. 
3 Dr G. E. R. Lloyd has suggested to me that the impossibility of looking for the good in a 
single form, though doubtless not to be directly related to the single Form of the Good in Rep., 
is nevertheless a further indication of the tendency towards increased flexibility in P.'s dialectical 
method. 
234 
Philebus 
most vehement pleasures are being taken into account. What has  
happened to those' consistent with health and virtue'? Well may Protarchus 
feel a residual twinge of doubt when Socrates has pronounced the 
inevitable verdict.' Well', he says cautiously,' so it appears at least from 
what has now been said' F6a9). 
The five possessions F6a-c). We have been told so far that the best life 
contains three elements: intelligence (including knowledge of any kind), 
pleasure (strictly limited in kind) and reality. To conclude, Socrates 
divides what he calls 'possessions'1 into five, on a scale of decreasing 
value. To determine the difference between the first two has been a 
standing puzzle with scholars for many years, intensified by an  
unsolved textual crux.2 There is also the old difficulty of the article-plus- 
adjective construction. Are 'the symmetrical', 'the beautiful' etc., here 
things of a certain character, the character they have, or Platonic 
Forms? The relevant passage may be approximately rendered thus 
F6a5-b3): 
Pleasure is not the first nor yet the second possession. The first is found 
somewhere in the region of measure, what is within measure (?? ???????) 
and appropriate^ and everything that must be thought to be of this sort —4 
The second is in the region of the duly proportioned (or symmetrical, 
?????????) and beautiful, what is complete in itself and adequate to its 
purpose^ and everything of that kind. 
It is hard to believe, with some scholars, that no distinction of value or 
pre-eminence between these two is intended. In determining it we must, 
I think, limit the fivefold assessment to good things within human life, 
in view of the word 'possession' as well as the tenor of the dialogue as a 
whole. This excludes the Forms themselves, though Plato will have in 
mind that goodness for man, inseparable as it is from due measure and 
1 ????? 66a 5. Cf. 19 c ?? ??? ?????????? ???????? ???????. 
2 Bury discusses the problem fully in his App. b, 164-78, with copious reference to previous 
views. One may add Rodier, Etudes 134-7, Jowett in, 544, Friedlander, PL in, 350 with notes, 
Hackforth, P?P 137 f. 
3 ?? ???????, a reminder that we are dealing with human life. Cf. p. 171 above. 
4 There follow the doubtful words, which may mean, if ????? is retained, that measure etc. 
'have taken on the nature of the eternal*. The passage is discussed by Gosling, Phil. i37f. 
5 ?????? ??? ????? v. Cf. 20 d, and for the distinction Bury 177. 
235 
Philebus 
moderation, mirrors that of the cosmos ordered by the divine Mind.1 
In spite of Rodier's criticism, there may be something in Zeller's  
suggestion that the first 'possession' represents participation in the Form 
Metron and the second its effects, the elements of moderation and the 
like in human life.2 But nothing is certain save that the two together 
exalt the primacy of measure, moderation, the right mean, on which 
Plato expatiated at greater length in the Politicus (pp. 169-72 above). 
Third in the list come intelligence and theoretical knowledge,3 on the 
human level one must assume (p. 238 below); fourth, practical  
knowledge and true beliefs; fifth, the pure pleasures enjoyed by the mind 
alone, though some are occasioned by the senses.4 Enigmatic to the last, 
Socrates ends with a tag from an Orphic theogony: 'But in the sixth 
generation cease the order of your song.' Does this mean ' Do not go 
beyond the fifth' or 'End with the sixth'? The former has been main- 
tained,5 nor does Plato say what the sixth would be. Nevertheless the 
Orphic poet did include the sixth,6 and the reference is probably to the 
necessary and temperate pleasures which have been admitted to the 
good life on sufferance. 
There follows the coup de grace and the dialogue ends. Victory in the 
contest for the title 'the good' goes to the mixed life of reason and 
pleasure together, but' reason is a thousand times more closely akin to 
it than pleasure'. 
1 Cf. Rep, 500c. The words ?? ?? ????? at 64C9, and iv ????????* ?? ??? ????* at 65b2, if 
significant at all, will not imply more than this. 
* Zeller 11.1. 874f., Rodier, Etudes 134-7. There are difficulties, as Z.'s long note shows, 
especially in agreeing with him that though no. 1 is not the Form itself, only in no. 2 do we get 
the 'projection into reality' ('Einbildung in die Wirklichkeit') of the Form. It is hard to see just 
what P. would have had in mind as a third stage between a Form and its instantiation in the 
world, unless (which I do not advocate) it is something like the 'tallness in us' of the PAo., 
interpreted in a somewhat Aristotelian way. (Cf. vol. iv, 354-6 and Archer-Hind on Tim. 50c.) 
3 vous ??? ???????, ??????? being, as often, interchangeable with ???????? (vol. iv, 265). 
4 These correspond to the pure pleasures earlier described. There is some doubt about the 
text, on which see Rodier, Etudes 132. On the order of merit Crombie writes with his usual 
illuminating good sense, though confessing himself puzzled as everyone must (EPD 1, 264 f.). 
I have felt some doubt over his point that intelligence would be placed third because responsible 
for the first two. One would expect an ?????? to be prior to its effects. However, 22 d may justify 
the supposition that P. had some such notion in his mind. 
5 E.g. by Jowett in, 535: 'The fifth [place is assigned] to pure pleasures; and here the Muse 
says "Enough"', and further on p. 545. 
6 See Guthrie, OGR 82 with n. 4. 
236 
Philebus 
The philosophy of the Philebus 
i. The Forms and the sensible world. Plato's two-world scheme, of 
Being and Becoming, is basically unchanged. This is indisputable, 
though the phraseology of Being may have infiltrated into the purlieus 
of Becoming. (See additional note on p. 232 above.) At 61 ei we have 
the familiar antithesis between 'the things that come to be and perish' 
and those that * neither come to be nor perish', which are the subjects 
of different kinds of knowledge. We have also met other unmistakable 
references to the unchanging, absolute realities. Philosophical dialectic 
is still concerned with these E7e~58a), and the pleasures of knowledge 
remain highly exclusive, 'by no means for the polloi, but for the very 
few' E2b). At 62a Justice serves as a typical example, though at 15a 
we have besides the moral and aesthetic Forms those of natural species. 
The eternal realities, however, are not the subject of the dialogue. 
Plato may be repetitious, but not so much so as simply to enlarge once 
again on the metaphysical theories of the Phaedo, Republic, Symposium 
and Phaedrus. They can now be taken for granted. Like the Politicus, 
the Philebus is concerned with life in the body, and how to live it as 
well and happily as possible. 'The good' in several places may recall 
what we have learned of the Form of the Good, but I think this would 
be misleading, even in the phrase 'the good itself, as at 61 a and 67a. 
'The good' is the goodness in human life, which, life being a mixture, 
consists not in the actual ingredients but in the correctness of the 
formula to which it is made up. If the proportions are right, the mixture 
will be good.1 
2. Theology: microcosm and macrocosm.2 As diairesis has been said to 
overshadow the nominal theme of the Sophist, so theology, the concept 
1 It has been objected that a mixture compounded strictly to rule could as well be a poison 
as a healthy drink. Plato could with consistency reply that it did indeed owe to its proportions 
the achievement of its telos and so was a good poison. Whether or not it ought to be administered 
to an insect-pest or a person, or neither, is a question which does not concern the dispenser. In 
the same way a captain does his job well if he brings his ship and passengers safe to land. For a 
particular passenger drowning might have been better, if a worse fate awaits him at his destination, 
but that is not the captain's business. (See Gorg. 5iie-i2a.) This is not inconsistent with  
disapproval of Gorgias's contention that a teacher of rhetoric is not responsible for the use of it made 
by his pupils (ib. 457b). Arts concerned with material production, health or safety are subordinate 
and morally neutral. It is in their use by a higher art that good or evil enters in; and rhetoric 
is an art (or pseudo-art) with a moral influence as strong as that of the statesmanship which it 
mimics. 
1 Something has been said about this already, on pp. 203, 213-16. 
237 
Philebus 
of a divine Mind penetrating the universe and all things in it, is basic to 
the Philebus. Only on that basis can Plato's theory of the place of 
pleasure in the good life be justified. Mind is the fourth kind of being, 
the cause of the combination of Limit with the Unlimited which ensures 
that the cosmos shall exhibit the supreme merits of order, right measure 
and proportion Coa-c). It is of course an extra-cosmic God, yet there 
is no real problem in reconciling this with the statement that the cause is 
something within the mixture B2di, 64c5), for the cosmos itself is 
living and intelligent and so, within it, is man, possessing fragments' 
(as the Pythagoreans said) of the divine Mind, timebound and restricted 
by association with the body. (As the Timaeus teaches, the Creator is 
not omnipotent, and can only impart his own nature to his creation as 
far as the irrational force of Necessity, or brute matter, allows.)  
Considered as possessions of mankind, mind and thought may be given 
only third place because the first two are concerned with the divine 
gift of duemeasure, proportion, the right mean which he shares with the 
cosmos. But universally speaking primacy belongs to Mind or Reason 
in its purity, 'King of heaven and earth', giver of beauty, symmetry 
and truth which are the marks of goodness. For full understanding, the 
Philebus must be read in the light of its sister-dialogue the Timaeus, 
which Plato may well have written first. The same applies to the 
Politicus, where the nature and importance of metron and the metrion 
are more fully and clearly explained, though without the theological 
overtones. (See pp. 169-73 above.) 
Conclusion. To be subjective in my turn, if Plato ever wrote a * weary' 
dialogue (p. 164) it is not the Politicus but the Philebus. This is not the 
fault of its main subject, the central importance of which he rightly 
emphasizes in the Laws. * Human nature involves above all things 
pleasures, pains and desires. Every mortal animal is so to speak hung 
up on them and kept dangling like a puppet.' 'When men consider 
legislation, practically the whole enquiry concerns pleasures and pains, 
in both communities and private individuals.'1 What tries the reader is 
a certain untidiness, and a lack of that precision which Plato himself 
singles out as the mark of real knowledge. Outstanding is his unqualified 
1 Laws 7326 and 636c!. For the language of 732? cf. what is said of the ???? at 644c 
238 
Philebus 
use of the word 'pleasure' to mean sometimes all that he most disliked 
in the popular notion of pleasure, and sometimes what Philebus and his 
like do not admit to be pleasures at all. He is at pains to show that the 
only pleasure worth having is the pleasure attendant on mental exercise 
—neither, that is, what others call pleasure nor he himself in other parts 
of the dialogue, both before and after his demonstration that they are 
falsely called pleasures. He admits' necessary' pleasures to the good life, 
which can only be the pleasures of moderate eating and drinking and of 
sex within marriage, on which the preservation of individual and race 
depends. (Cf. 35e.) But one feels he would have been happier if these 
ends could have been achieved without pleasure. Eating, drinking and 
sex, after all, provide the measureless, the 'greatest and most intense' 
pleasures which cannot be admitted at all.1 More generally, much of the 
dialogue repeats obscurely what has been more fully and clearly 
explained elsewhere. 
The whole question at issue is really settled from the beginning. It is 
a statement of belief rather than a genuine argument. Socrates makes 
a series of dogmatic pronouncements which go unchallenged and which, 
once granted, make the victory of philosophy over pleasure a foregone 
conclusion. For a defender of hedonism it is absurd that Protarchus 
should agree to them as he does (with the temporary exception of the 
existence of false pleasures). This, I surmise, is why the hard-line 
hedonist Philebus had to be replaced by a pliable youth who only 
thought he was a hedonist. In the Protagoras Socrates himself  
demonstrated how the art of measurement was reconcilable with hedonism, 
indeed essential to it; but not here. The whole argument is based on 
premises both intellectual and moral which a Callicles or Philebus would 
deny. Perhaps this is a reason for its obscurity: it is not genuine 
argument but a fasade for dogma. 
Another reason may be that it is a compromise. Plato retains his 
conviction that the philosophic life is the best, but knows that very few 
can practise it at all, and those not all the time. He is in the ambiguous 
position so strikingly brought out by Aristotle, who shared it. Perfect 
human happiness and self-sufficiency, he says in the Ethics (bk 10 ch. 7), 
1 Cf. Laws 782e-83a: The appetites for food, drink and sex are all distempers (????????), 
to be held in check by fear, laws and sound reasoning, together with the moderating influence of 
the arts. 
239 
Philebus 
lie in intellectual activity, but to spend a whole life in it would be 
superhuman. A man will live it not qua human being but in so far as he 
has in him something divine. In the very next sentence we are exhorted 
not, as the poets recommend, * being mortal to think mortal thoughts', 
but to aim at immortality as far as we can and live according to the best 
in us, which would even seem to be each one of us. The dilemma is 
patent. In the same breath Aristotle can speak of the life of reason as too 
high for men and exhort us to pursue it as really and truly our own. 
Man's position in the world is unique, because like no other creature he 
houses reason, which is divine, in a mortal body. All Plato's inclinations 
were towards cultivation of the divine part, and when, as in these later 
dialogues, he determined to allow full weight to the necessities of our 
incarnate state, the resulting tension led to an unevenness, and even 
downright inconsistency, disturbing to a reader but rendered more 
comprehensible by Aristotle's prosaic assessment of their common 
position.1 
1 P.'s own clearest statement of it is at Tim. 9ob-d. Aristotle's ??* ???? ????????? ??????^??? 
is practically a quotation of 90c 2-3, ???* ???? ?* ?? ????????? ????????? ????? ????????* 
?????????. In Phaedo the word 'man' (???????*) is reserved for the compound of body and soul 
(95 c6, 76c 11), but in Ale. I we have ? ???? ????? ???????? A30c). 
24O 
VI 
TIMAEUS AND CRITIAS 
Introduction 
It must be admitted that few books created so much intellectual evil as 
the Timaeus; the only one which created a greater perversion of thought 
in the Christian world was the revelation of St. John the Divine. 
G. Sarton 
Our illustrator of the atomic model [in a school text-book of physics] 
would have done well to make a careful study of Plato before producing 
his particular illustration. W. Heisenberg 
The influence of the Timaeus down to the Renaissance was enormous, 
and interest in it has continued unabated, if from different motives, to 
the present day. One of its most perceptive commentators, Th. H. 
Martin, called it 'the most quoted and least understood' of Plato's 
dialogues. Plato's younger contemporaries were already disputing its 
meaning. Aristotle cites it more often than any other dialogue, and 
thought it worth while to write an epitome. Whatever he may have 
known of any * unwritten doctrines', he took the Timaeus as a serious 
exposition of Plato's own philosophy and science. The first commentary 
was written by Crantor, a pupil of Xenocrates, about the end of the 
fourth century B.C., and the Hellenistic and Roman periods saw  
contributions by Stoics like Posidonius and Panaetius, Cicero (who translated 
it), Plutarch (in his On the Generation of Soul in the Timaeus and 
Platonic Questions) and others. The Christian world received a portion 
of it in the fifth century in the Latin version of Chalcidius (to 53 c),1 
through which alone it was known until the twelfth. Klibansky's  
comment might surprise some critics of Plato, that in an age marked by an 
attitude of contempt for the world, it was he who kept alive 'the 
Hellenic appreciation of the rational beauty of the universe'. For that 
reason the Timaeus deeply influenced the philosophers of the Renais- 
1 The date of Chalcidius is controversial, but Waszink, the editor of his Timaeus, thinks he 
wrote it a little after 400. 
17 
24I 
GHG 
Timaeus and Cntias 
sance. Later, as scientists increasingly followed Bacon's advice to 
detach physical science from final causes, talk of which, 'like barnacles 
on a boat, holds up the voyage of science', its study was left increasingly 
in the hands of historians and scholars. Most recently however, in spite 
of its lack of experimental method, its geometrical theory of the world 
has come into its own again as evidence of a brilliant natural insight into 
the structure of matter. Whitehead had already written in 1929 that 
* Newton would have been surprised at the modern theory and the 
dissolution of the quanta into vibrations: Plato would have expected it', 
when authorities like Jeans and Singer were still describing the influence 
of the Timaeus as a scientific disaster and a degradation of knowledge. 
Now we have Popper's claim that the geometrical theory of the world's 
structure, which makes its first appearance in Plato, has been the basis 
of modern cosmology from Copernicus and Kepler through Newton 
to Einstein, and Heisenberg's opinion that the tendency of modern 
physics brings it closer to the Timaeus than to Democritus. Yet 
Democritus had for long been hailed as the true precursor of scientific 
atomism, partly perhaps because, as Bacon said in his praise, he 
assigned the causes of things to material necessity sine intermixtione 
causarum finalium.1 
However, our present task is not to consider later developments, but 
(to quote Rivaud) 'de nous mettre en presence du texte meme et de 
tenter de Pentendre'. So after a few remarks on date and personalities, 
we shall start by considering the general purpose and framework of the 
Timaeus-Critias ensemble. 
1 Rivaud has a brief summary of the influence of Tim. on pp. 3-5 of his introduction. For the 
Middle Ages see Klibansky, Continuity 28 f. (quoted by H. D. P. Lee, Tim. 23 f.). Aristotle's 
epitome is mentioned by Simpl. Cael. 379.16 Heib., Crantor's exegesis by Procl. Tim. 1.76.1 
Diehl. For Cicero's translation, of which fragments are preserved, see F. Pini, M.T. Ciceronis 
Timaeus A965) and R. Giornini, 'Osservaz. sul testo del Timeo ciceroniano' in Riv. di Cult. 
Class, e Med. 1969. Bacon's De augm. sc. bk 3 ch. 4 (Spedding, Ellis ed., 1, 568-70), quoted here 
in translation, is a powerful indictment of the introduction of final causes into physics. For reff. 
to modern scientific opinion see Friedlander, PL 1, 264ff. with notes. Add Sarton in Isis, 1952,57; 
Heisenberg, Physicist's Conception of Nature 60f. and 'P.'s Vorstellungen von den kleinsten 
Bausteinen der Materie und die Elementarteilchen der modernen Physik* in Im Umkreis der Kunst 
(Festschr. Pretorius); Singer, Short Hist. ofSc. Ideas A959), 40; Popper, Con}, andRef. 88 n. 45, 
89-93. But Democritus had a stout defender in Schrodinger. See his little book Nature and the 
Greeks A954), esp. p. 82: 'Democritus was intensely interested in geometry, not as a mere 
enthusiast like Plato; he was a geometer of distinction.' 
242 
Introduction 
Date. It should perhaps be said first that Morton and Winspear claim to 
have established, by stylometric tests with a computer, that the first 
300 sentences of the Timaeus are by Speusippus, not Plato.1 
Until 1953, the Timaeus and its sequel the Critias were universally 
believed to be, with the possible exception of the Philebus, the latest of 
Plato's works except the Laws. In that year G. E. L. Owen published 
his now famous article designed to show that on the contrary it  
belonged to the middle group of Republic and Phaedo and preceded the 
* critical' group. This has aroused a great deal of comment, mostly but 
by no means entirely adverse, and there is no point in adding to it,2 but 
the importance of Owen's thesis must not be underestimated. It carries 
with it a radical change in previously accepted ideas of Plato's  
development and a re-assessment of the philosophical basis of the critical 
dialogues (including the Philebus\ which in Owen's judgement 'gain 
in philosophical power and interest when they are read as following and 
not paving the way for the Timaeus' {SPM 313). So read, they can be 
interpreted as teaching a more sophisticated metaphysic based on 
renunciation of the doctrine of paradigmatic Forms and the opposition 
between Being and Becoming. Since I have tried to show at several 
points that these doctrines still make their appearance in Theaetetus, 
Sophist, Politicus and Philebus, I naturally find this view difficult to 
accept, but Owen's arguments must be read for themselves before 
anyone decides which course will lead him to the true mind of Plato. 
1 Gk to the C. 13. They do not say to what point in the dialogue this brings us, and I confess 
to not having counted up the sentences to find out. 
2 Owen, 'The Place of the Tim. in Plato's Dialogues', CQ 1953, repr. in SPM 313-38. 
(Before Owen wrote, the arguments in favour of the late date were summarized in Rivaud's 
edition, pp. 21-3.) The first notes of misgiving were sounded by Vlastos in 1954 (see now 
SPM 245 n. 3 and 247 n. 4) and by Field in a communication to the Classical Association 
(Proceedings, 1954, 52). Cherniss published a long rebuttal in AJP 1957, supported by another 
in JHS 1957 A). See also Skemp, P.'s Statesman 237-9, A. and P. in Mid-Fourth Cent. 201 f., 
TMPLD 68, de Vogel, Philosophia Pt 1 A964), 190 n. 2 and 237f. and Cherry in Apeiron 1967. 
Among brief or occasional criticisms that I have seen are D. Tarrant, CQ 1955, 224; Runciman, 
SPM 152 (i960); Reiche, Empedocles* Mixture 87f.; Herter, Rh. Mus. 1957, 347 n. 66 and 
Palingenesia ? ? A969), 117 n. 35; M. A. Stewart, PQ 1971, 172. 
In favour of Owen's re-dating have been Ryle, P.'s P., ch. 7 and elsewhere, and Ency. Phil. 
vi, 320; T. M. Robinson in AJP 1967, p. 57 ?. ? and cf. Anton and Kustas, Essays, where on 
p. 353 n. 31 he mentions (with full reff.) John Gould, D. W. Hamlyn, D. A. Rees and C. Strang 
as also accepting it. 
On the computer-based stylometric results of Cox and Brandwood (J. ofR. Statist. Soc. 1959, 
195-200) see Robinson (last ref.) and for a different estimate Stewart, I.e. 
243 
17-2 
Timaeus and Critias 
About the relation of the Timaeus to the Philebus one cannot be 
positive. I have been strongly tempted to put it earlier on the ground 
that doctrines elaborated at length in the Timaeus are briefly assumed in 
the Philebus, giving Ryle very good reason for saying that it 'echoes' 
the Timaeus. But for one who still inclines to the traditional date of the 
latter, there seems no possible reason why Plato should have broken 
off his grand trilogy (p. 246 below) to write a dialogue like the Philebus. 
That he should have abandoned it for the Laws, as a better treatment of 
the same subject, namely the best possible human society, is  
understandable, but for the Philebus—no. 
Characters. Besides Socrates these are Timaeus, Critias and Hermo- 
crates. The main part of the wos k is a continuous lecture by Timaeus, 
of whom we have no authentic information other than what is given in 
the dialogue, namely that he is a prominent citizen of Locri who  
combines statesmanship with philosophy and is especially well versed in 
astronomy and cosmology B0a, 27a). He may or may not be a 
historical figure (cf. Martin, Timee 1, 50), but at any rate his Italian 
origin implies that his philosophical affiliations are with the West. 
Hints of the age of Critias make it appear that he is not the oligarch and 
second cousin of Plato, but his grandfather. Hermocrates is identifiable 
with the Syracusan general, highly praised by Thucydides, who foresaw 
and then defeated the Athenian aggression against Sicily. He was the 
father-in-law of Dionysius II.1 
Framework and purpose 
Socrates opens the Timaeus by repeating the main heads of a talk which 
he had given the day before to the three now present and one other,2 on 
the best form of political association. This recalls many of the political 
and social provisions of the Republic. At his request, the others are now 
to bring the planned city to life and show it in action, and especially in 
the conduct of war and international relations generally. As experienced 
1 For further details of all these see Cornford, PC 1-3. Popper cites 'Plato's eulogy of an 
enemy of Athens like Hermocrates' as part-evidence of his hostility to contemporary Athens 
(OS 1, 311). Similarly Rivaud, 77m. 15. On Critias I have given what is now the general opinion, 
instanced by (besides Cornford) Taylor, Comm. 24; but Vidal-Naquet thought it unnecessary to 
insist on such chronological niceties in Plato (REG 1964, 420 n. 3). 
2 There is no hint of the identity of this unnamed person, said to have been taken ill, nor of 
P.'s motive in mentioning him. 
244 
Framework and purpose 
politicians they are better fitted for this than a theorizer like himself. 
They then unfold their plan. Critias will tell the story, preserved in 
Egyptian tradition, of the defeat of the aggressive power of Atlantis by 
Athenians of 9,000 years ago, whose institutions, by a providential 
coincidence, closely resembled those of the state imagined by Socrates. 
He will take this Athens of a bygone heroic age to have actually been 
that ideal state, thus illustrating its successful conduct of a war, and the 
others will follow him in a programme designed to satisfy Socrates to 
the full. 
The self-criticism of Socrates as an impractical theorist is striking. 
For the Socrates of the Republic it was not important that his model 
state should ever be realized in practice; it remained as an ideal {para- 
deigmd) to guide men's footsteps towards justice and the right D72 d- 
73 b). But now he wants to know how it would fare in the rough and 
tumble of clashes with other states. This, surely, is the purpose of the 
opening reminder of some of the Republic's measures. Plato is telling us 
explicitly that in the years since he wrote it his interests have veered 
from an idealistic view of society towards practical policy, as the 
Politicus has already shown.1 Another lesson of the introduction is that 
Plato's talent for casting his ideas in dramatic form, the creative  
imagination that puts him in the front rank of poets and story-tellers as well as 
philosophers, is undiminished. The portrayal of Atlantis, when it comes, 
is so vivid that many scholars and geographers have supposed that it 
really existed. Most of it comes, however, not in the Timaeus but in the 
Critias. The opening of the Timaeus is an introduction not to that 
1 For the Pol. see pp. 171, 184, 187 above. Dramatically, P. differentiates S.'s exposition 
from the Rep. itself by making the narration take place on the day before the Panathenaea instead 
of the day after the Bendideia (??????? ???? .. .). The additional point raised by Rivaud (p. 3), 
Ryle (P.*s P. 230) and others, that the characters of 77m. are not those of Rep., is strictly speaking 
irrelevant. However unrealistically, the Rep. is in form a continuous narrative by S. of his 
conversation in the house of Cephalus on the previous evening, and nothing is said of the audience 
to whom he is relating it. Of course, against the strange view of 77m. as a * continuation* of Rep., 
Raeder's objection on grounds of form is sound enough (P.*s Ph. Entw. 195), and to regard the 
characters in 77m. as the silent auditors of Rep., though logically possible, is implausible. The 
passage has led to some curious theories, e.g. Rohde's invention {Psyche 477 f.) of an earlier, 
incomplete edition of Rep., because only some of its tenets are mentioned and Timaeus 
agrees that S. has recalled in summary (?? ?????????) the whole of 'yesterday's' speech— 
surely a minimal dramatic licence. The fact remains that he does list unmistakable excerpts 
from Rep., and the motive for this I have suggested in the text. Only a few reminders were 
necessary. 
245 
Timaeus and Critias 
dialogue but to the Critias, or rather to the whole trilogy1 which Plato 
planned but for some reason abandoned after writing a small part of the 
Critias, so that we can never see it in its true proportions. 
It is essential to see the Timaeus in this perspective. Plato has not 
abandoned human affairs for cosmogony and natural science. His  
purpose is to place man in his setting in the world and draw out the  
implications for human life and aims. The theme to be elaborated in detail is 
that taken for granted in the Philebus and already hinted at in the 
Republic, namely the close relations between microcosm and  
macrocosm. ' Familiarity with the divine and orderly makes the philosopher 
divine and orderly so far as a man may be ' {Rep. 500c). In the Timaeus 
this becomes: 'By learning to know and acquiring the power to  
compute them rightly according to nature, we may reproduce the steadfast 
revolutions of the universe and reduce to settled order the wandering 
motions in ourselves.' For Plato this is now the essential prelude to his 
new vision of the good society in action. Similarly in the Laws (bk 10), 
which appears to have replaced the rest of the trilogy, a conviction that 
the universe is rationally and divinely governed is the prerequisite for a 
moral human life. In the Phaedo Socrates abandoned natural philosophy 
altogether because its exponents asked only how, not why, things  
happened. This was true of the real Socrates, and at that time Plato  
followed his lead.2 In the Philebus he says E8C-59C) that most arts are 
concerned only with beliefs (doxai). Even the students of nature confine 
themselves to this world, how it came to be and what goes on in it; but 
all this has nothing to do with what is real and unchanging, nor can its 
study by itself lead to knowledge of the truth. This contrast between 
changeless reality and changing phenomena, knowledge and belief, is 
the starting-point of Timaeus's discourse. Now however Plato has 
decided that it will not do simply to point out the physicists' error and 
1 That a third dialogue Hermocrates was planned is clear not only from the promise of Critias 
at 27d that all will take part, dividing the work between them, but from Cr. 108a: 'when it is 
H.'s turn to speak*. What it was to contain we are not told, but Cornford's guess is plausible 
{PC 7f.). Critias's story ends with one of those natural catastrophes mentioned in Pol. B70c-d) 
and here at 22 c-d—in this case earthquake and flood—after which civilization has to start again 
from the rude beginnings of a few illiterate survivors in the hills. Just such an extinction and 
rebirth of culture are described in Laws 3 and 4, which may reasonably be supposed to replace 
what H. would have said when he took up the tale from Critias. 
2 For S.'s own attitude to natural philosophy see vol. in, 421-5, and for his belief that the 
world is divinely go\erned, lb. 442. 
246 
Framework and purpose 
leave the subject. He must challenge them on their own ground by 
constructing his own De rerum natura, which on the level of physical 
causes will incorporate much earlier work but will relegate them to their 
proper place as secondary, necessary indeed but only auxiliary to the 
creation of a rational order modelled on the changeless Forms. Only in 
the light of the final cause can the physis of things be really understood. 
Like Anaxagoras, Plato starts from the axiom 'All things were in  
confusion, then Mind came and set them in order', but unlike Anaxagoras 
the Ionian he will continue to maintain it as the premise on which all 
conclusions about the physical universe must depend. It has been said 
that ' The basic question of philosophy is the question concerning the 
relationship between thought and being, between spirit and nature— 
which came first.'1 The Timaeus is Plato's full and definitive answer to 
that question, and this must never be forgotten when we turn to the 
self-confessed obscurities and difficulties of his detailed account of the 
physical world. 
One further point. In trying to interpret his thought, I shall not feel 
bound to follow the order of Timaeus's exposition. Plato has retained 
one feature of his conversational manner: even in a continuous account 
he still aims at the effect of an unrehearsed talk. Themes intertwine, one 
being pursued until Timaeus remembers something that he ought to 
have said earlier and goes back on his tracks to pick it up. (See  
Appendix, pp. 39if.) 
Atlantis {Tim. 2od-25d, Cr. io8e-2ic) 
Few commentators spend much time on the marvellous story of 
Atlantis, and a general historian must certainly resist the temptation to 
do so. A bibliography of 1926 listed 1,700 items on the subject, and 
it would need little research to add another 50 or more.2 There are 
those who believe, with Martin (p. 332), that 'elle appartient a un autre 
monde, qui n'est pas dans la domaine de l'espace, mais dans celui de la 
pensee', and those who fight hard to give it a terrestrial locality. It 
1 I. M. Bochenski, Dogmatic Principles of Soviet Phil. I, 13. 
2 No attempt at a bibliography will be made here. Martin (?tudes 1, 257-332) deals  
comprehensively with the literature up to 1840. See also Rivaud's ed., 27-32. James BramwelPs Lost 
Atlantis A937) gives a highly readable accounts of facts, theories and fancies from Crantor to the 
present century. 
247 
Timaeus and Critias 
has been sought from the Arctic Ocean to North Africa, from America 
to Ceylon. Interest in its historical existence has now been given fresh 
stimulus by the discovery of Minoan remains on the volcanic islands of 
Santorin (Thera and Therasia), destroyed about 1500 B.C. by an  
enormous eruption. According to Plato Atlantis was swallowed by  
earthquake and floods in a single day and night, and the Greek seismologist 
Galanopoulos believes that the metropolis of Atlantis was on Thera  
itself. Others, notably J. V. Luce, identify it with Minoan Crete, which  
certainly suffered severely from the effects of the eruption.1 It is certainly 
true that if the Atlantis story reflects a historical disaster, Thera provides 
the only known example on a sufficient scale to be taken seriously. 
It must be remembered that Plato is our only authority for the story,2 
which he ascribes to Solon, who learned it from Egyptian priests and 
told it to his contemporary Critias, who at the age of ninety repeated it 
in the presence of his grandson the narrator in Timaeus and Critias, who 
was then about ten B1a). Now Plato tells us two things: A) Atlantis 
was a huge island, or rather continent, lying in the Atlantic Ocean just 
outside3 the Straits of Gibraltar.4 That a submerged land in such a 
situation is a geographical impossibility has been made a reason for 
finding it elsewhere, but is of course irrelevant if one regards the whole 
story as the product of Plato's imagination. B) Its destruction took 
place 9,000 years before Solon heard of it {Tim. 23 c, Cr. io8e). On the 
Santorin theory it should be about 900. Galanopoulos has attributed 
this to a mistranslation by Solon of the not very similar Egyptian 
symbols for 1,000 and ioo;5 but the mistake would have been the 
1 The theory that Atlantis is a memory of the Minoan civilization goes back to ?. ?. Frost 
mJHS 1913. See now A. G. Galanopoulos and E. Bacon, Atlantis, the Truth behind the Legend 
A969), and J. V. Luce, The End of Atlantis: New Light on an Old Legend A969). 
2 It is true, and I have not seen it mentioned, that Plutarch in his life of Solon (ch. 31) says 
that he left the writing of the story unfinished 'not, as Plato says B1 c), through other  
preoccupations but rather through old age'. This might suggest a second source, but in context sounds 
more like a personal surmise on the part of Plutarch's own authority. 
3 'Before the entrance' (???? ??? ???????? 24a) I take to mean that it was at no great distance, 
but the volcanic Azores have a better geographical claim to be the remains of Atlantis than any 
spot within the Mediterranean. They were first proposed in 1787 (Bramwell 137). 
4 As to the name, Herodotus called the sea beyond the Pillars of Heracles '???????? A.202). 
Plato B4 c) calls it ?? ?????????? ????????. The Pillars were known also as ???????? '?????????? 
(Eur. Hipp. 3). All these names were presumably taken from the 6pos ?? ?????? "?????, the 
dwellers around which are called "???????? (Hdt. 4.184). 
5 Besides G.'s book, cf. The Times for 13 Feb. 1962: 'Professor Galanopoulos told your 
correspondent: "Solon, in translating Egyptian texts, systematically mistook the symbol for 
248 
Atlantis 
Egyptians' own. Solon did not try to decipher hieroglyphics: he was 
told the story by the priests B3dff.). A mere 900 years would also be a 
sad anti-climax immediately after the impressive testimony to the 
extreme, antediluvian antiquity of Egyptian temple records B2b-23b). 
Moreover at 111 a Critias says that many great floods have occurred in 
the 9,000 years since the time of which he is speaking. 
But the most persuasive case for identifying Atlantis with Minoan 
Crete was presented after Galanopoulos and Luce by the Cretan 
archaeologist Nicholas Platon in ch. 35 of his book on Zakros A971). 
Undue importance, he says, has been given to the fact that Plato places 
it beyond the Straits of Gibraltar (whereas I am inclined to think that 
this fact has been too much ignored). 'According to the earlier cosmo- 
graphic conceptions' it was in the centre of the Mediterranean that 
Atlas supported the vault of heaven. Unfortunately he gives no  
references for this, and I have failed to find any hard evidence. In RE 1, 
2127, Wernicke certainly demonstrated that the earliest associations of 
Atlas the Titan were with Arcadia, and adds 'And so [or 'therefore', 
also] in Arcadia he supports the sky.' It follows that this belief will have 
been held by the pre-Greek dwellers in Arcadia. 'How natural it was 
for them . . .', he exclaims, and adds that they equated Atlas with the 
sky-god. Wernicke's article shows signs of the mythological theories 
of its time A896), not all of which are acceptable today, and in classical 
times Atlas's task was certainly associated with the North African 
mountain. However that may be, the strength of Platon's case does not 
lie there. The historical coincidences which he reveals between Plato's 
description of Atlantis and the physical equipment, layout and  
institutions of Minoan cities as elucidated by excavation are both fascinating 
and impressive. Here, I should say, lies the strongest argument that 
besides Plato's own fantasy, old memories of the Minoan empire, 
perhaps preserved in Egypt as he says, have gone to build up his picture 
of Atlantis. 
Though Bacon in his New Atlantis depicted the islanders as bellicose 
and imperialistic, and Ruskin told the tale of its destruction to the 
manufacturers of Bradford as a warning against materialism, Atlantis 
???' for that of ????'".' (Would P. perhaps have replied in words like those he uses about 
Egyptian art in the Laws F56c), ??? ?? lrco% ??????? ????????? ???' ??????) 
249 
Timaeus and Critias 
has inspired generations of writers as an ideal, another Eden, Isles of the 
Blest, Paradise or Utopia. This is curious, because in Plato's story the 
Atlanteans are the imperialist villains, the Athenians the heroes who by 
'going it alone' and repulsing them saved their fellow-Greeks and all 
the Mediterranean peoples from enslavement. Their success was due to 
the invincible spirit of unity resulting from an education and  
institutions miraculously like those of the 'best society' described by Socrates 
on the previous day. The Atlanteans on the other hand, though once 
virtuous, had degenerated into a covetous and power-seeking lot, ripe 
for divine judgement. Reading the last extant words of the Critias one 
does not need much imagination to see in them Plato's strictures on the 
Athenians of his own day, so different from those who beat back the 
Persian hordes at Marathon and Salamis.1 
Whether or not Plato's tale of Solon is all imaginary, or dimly 
reflects a folk-memory of some past event or distant land, Atlantis has 
taken on a life of its own of which neither time nor scepticism can rob 
it. Perhaps John Masefield had the last word: 
The Atlanteans have not died; 
Immortal things still give us dream. 
The 'probable account9 
When Protagoras says ? shall tell you no longer a mythos but a logos' 
(Prot. 324 d), his meaning is clear. So far he has cast his view of human 
nature into the form of a fictional narrative: the rest he will impart as a 
straightforward statement of fact. Each conveys truth in its own way. 
In the Timaeus Plato does not even distinguish them. Timaeus calls his 
discourse a mythos or a logos indifferently, though more often the latter,2 
and regularly with the epithet 'probable' or 'likely' (eikds). The first 
problem facing an interpreter is how far the account is intended as 
serious philosophy or science. In the nature of the case, the text cannot 
decide this for us, and opinions are inevitably coloured by the indi- 
1 The chief thesis of Vidal-Naquet's interesting and thought-provoking article in REG 1964 
is that Athens' conquest of Atlantis was a conquest of herself. Athens and Atlantis represent two 
sides of the same city. Primitive Athens is the Athens of the land, of Athena and the olive, 
whereas Atlantis, founded by Poseidon, is the imperialistic sea-power which she became. 
2 See Vlastos's figures and reff. in SPMy 382 with notes. Note that at 56b he claims to speak 
???? ??? ????? ????? ??? ???? ??? ??????. 
25O 
The ?probable account* 
viduaPs impression of Plato as a whole, which in its turn will almost 
certainly be influenced by his own outlook on the world.1 I claim no 
exemption for what I shall say in subsequent sections, but the text must 
have first consideration. 
Introducing his theme, Timaeus says Bc>c-d) that one cannot hope 
to give a completely consistent and precise account of such subjects as 
gods and the origin of the universe, but must be content with a probable 
mythos. At 48d he speaks of 'maintaining what we said at first, the 
force of a probable logos', and at 47c of 'holding fast to what is 
probable'. By contrast, the poets have spoken of the gods 'without 
probable and necessary demonstrations',2 and our trust in them can 
only be based on their divine descent Dod-e). Whether or not this 
amounts to a dismissal of their claims,3 it is clear that such methods are 
not going to be followed here. At 59 c, after speaking of the composition 
of metals (hardly a mythological theme, as Vlastos remarked), he calls 
this sort of analysis 'pursuing the method of probable mythoi\ and 
adds:' When a man for relaxation puts aside logoi about what exists for 
ever and gets an innocent pleasure from the probable logoi of becoming, 
he will add a reasonable and sagacious recreation* to his life.' Here is the 
clue to an assessment of the 'probable logos9. Only of being can there 
be certain knowledge: of the natural world, as a world of becoming, we 
can only have belief. This is the basic distinction which Timaeus laid 
down at the beginning Byd). But even beliefs can be 'firm and true' 
C7b), though perhaps we cannot be sure of it. 'That what has been 
said about the soul is the truth, we could only assert if God confirmed 
it; but that it is probable we must venture to say now, and the more so 
as investigation proceeds' G2d). 
1 Cf. Taran's remarks at the conclusion of his article on ' The Creation Myth in P.'s Tim. * 
(Anton and Kustas, Essays 392). The point 'cannot be settled by discussion, for what is in 
question is how one reads Plato, how one conceives Plato's role as a writer and as a thinker'; 
but the issues 'can be discussed and can at least be clarified by the use of argument'. 
2 40c Contrast 53d: the geometrical structure of the elementary particles is given ???? ??? 
???' ??????? ?????? ?????. Contexts like this tell strongly against Howald's assimilation of ????? 
in Tim. to the rhetorical ????? condemned by P. in other dialogues (Hermes 1922, 70f.). 
3 The difference is one of subject: cosmic and astral beings, or the denizens of Olympus. The 
passage is generally taken to be wholly ironical, but cf. Guthrie, OGR 240 f. 
4 Paidia. On this word cf. vol. iv, 61. At 69 a (see next page) this same study of nature has 
become the indispensable preliminary to the knowledge of divine causes which is the primary 
objective. Nor should we overlook the fact that the manuscripts are divided between ?????? and 
???????. See Burnet's app. cr. 
251 
Timaeus and Critias 
The progress of Plato's thought is subtle. He has abandoned nothing 
of his conception of reality as incorporeal and supra-sensible, or of 
divine purpose as the ultimate cause of everything being as it is; yet his 
present view of science as advancing indefinitely through investigation 
of phenomena without ever reaching unquestionable truth has more in 
common with the twentieth-century than the Aristotelian conception 
of its progress. In the Phaedo, though the senses could stimulate the 
mind to search for the Forms, the instability of the world of change 
made it waste of time to study it in detail. Now it is only through that 
study that we can hope to attain the knowledge of divine and changeless 
reality, because even the divine Craftsman, in making the world as 
good as possible, availed himself of given material and secondary 
causes. 
Therefore we must distinguish two types of cause, one necessary, the other 
divine. Divine causes we must ever be seeking in order to secure a life as 
blessed as our nature admits, and the necessary for the sake of the divine, 
reckoning that without them we can never apprehend in isolation those other 
things on which our mind is set, nor receive nor in any way have part in 
them. F8e-69a) 
The reasons why only a likely account of the natural world is  
possible are twofold, objective or ontological, and subjective, (i) An 
explanation must be conformable to its subject. The world is only a 
changing likeness (eikon) of an unchanging model (paradeigma\  
therefore its description can only be provisional and likely (eikos), not final 
and immutable like a logos of the model. B) Neither speaker nor listeners 
can transcend the limitations of human nature. (See 2c>b-d.) 
The upshot is that Plato intends his account of the natural world to 
be as accurate as possible within the limits imposed by the subject- 
matter and man's powers of understanding.1 Much of the contents bears 
this out. The human physiology and pathology, for instance, reflect the 
latest opinions of the Sicilian medical school. Yet this does not settle 
every question. What of the liver as literally a mirror of the mind (made 
'solid, smooth and bright' for the purpose), with the function of induc- 
1 Note that in introducing the geometrical structure of matter Timaeus emphasizes that his 
logos is intended for experts. It is unfamiliar, he says,' but since you are versed in the branches of 
learning which I must employ to demonstrate my thesis, you will follow me' E3c). 
252 
The 'probable account* 
ing prophetic dreams G1 a-d)? What of the mixing-bowl in which God 
blends the ingredients of the souls of the world and of men D1 d)? 
Plato believed in transmigration, but did he believe that at first only men 
were created, and women originated from inferior males at a second 
birth? Hardly, yet this statement is accompanied by a solemnly scientific 
account of the physiology of sexual reproduction. On a higher level, 
some regard the Creator himself as mythical, identifying him with his 
model or the World-Soul—which according to the 'probable account' 
he brought into being—or both. How far the Timaeus is intended to be 
mythical, and what exactly 'mythical' means, will never be settled now 
by argument, if indeed Plato could have settled it himself. Jowett 
thought he could not,1 and it is remarkable that Plato's own pupils, 
Aristotle and Xenocrates, differed over whether the temporal creation 
of the world was intended to be taken literally. 
Maker) Model and Material 
As Plato presents it, his cosmogony demands three, or perhaps four, 
ultimates: the Maker, his eternal Model, and the unformed material on 
which he worked, and which was pervaded by a restless, irrational 
motion from no other cause than necessity or chance (Ananke). Let us 
take these one by one. 
The Maker. In the Republic Plato calls the maker of the heavenly bodies 
their demiourgos, and this word is used several times in the Timaeus. 
Consequently he is now usually known as the Demiurge, though Plato 
more frequently calls him God (that is, usually theos with the definite 
article, to distinguish him from the many derivative gods), and also 
Father and begetter.2 In the latter metaphor his raw material is  
compared to a mother, according to the usual Greek beliefs about parentage. 
1 Dialogues, vol. in, 698: 'We cannot tell (nor could Plato himself have told) where the figure 
or myth ends and the philosophical truth begins. * Not surprisingly when Plato, like his  
Protagoras, used myth to convey philosophical truth. 
* Rep. 530a, and cf. 507c, Soph, 265 c, Pol, 270; ?????????? in Tim, 41a, 42?, 68e, 69c (and 
verb ??????????? 37 c); with ????? at 41a, ? ???? ?? ??? ???????* 41a, ??????? ??? ????? 28 c 
(cf. Phil. 27 a: ?? ?????????? and ?? ?????????? differ only in name), ? ???????* ????? 37 c 
? ???$ is the most frequent title C0a, b, d, 31b, 32b, 34a, 55c, 56c, 69b, 73b). It is going too 
far to say with Cherniss {AC ? A 608) that sing, and pi. are used 'practically without  
discrimination'. 
253 
Timaeus and Critias 
However, the choice of Demiurge, meaning craftsman or technician, is 
sound. The spirit of Socrates still lives, with his endless talk of'  
shoemakers, carpenters and smiths', and the word reminds us that a  
craftsman works in a given material and to a pattern or form, either before 
his eyes or reflected in his mind.1 Similarly the Maker of this world is 
not omnipotent, but does the best he can with an already existing stuff, 
and creates the physical cosmos after the model of the eternal Realities. 
The metaphysic of the Timaeus is not monistic in the sense that One 
Being is primary and all else derived from Him. In the Philebus we have 
seen it put plainly and succinctly B3c-d): for anything in this world to 
exist, there must first be, as its constituents B7 a 11), an Unlimited 
element and the principle of Limit or proportionate measure to be 
imposed on it, and besides these a Cause to effect their union. (Cf. 30c.) 
The Cause does not create these two, but only blends them to make the 
concrete object. This supreme Cause in the Philebus is Reason, more 
fully personified in the Timaeus^ where among the Demiurge's many 
titles is 'Best of causes'.2 
It is the lesson of the Gorgias all over again E036-504^: 
The accomplished speaker who aims at the best result will not choose his 
words at random, but with his eye on something—just as all other craftsmen, 
each with an eye to his own work, do not pick out at random the materials 
they bring to it, but so that what they are making shall have a certain form. 
Look at artists,3 builders, shipwrights or followers of any other craft, how 
each of them imparts a certain arrangement to what he is working on, and 
makes one part fit and harmonize with another until he has constructed the 
whole as a thing of system and order. 
The goodness of anything, Socrates continues, whether house, ship, or 
the human body and soul, depends on whether it manifests order 
(kosmos). The world itself is a kosmos, as we know, and the action of the 
1 Cf. Skemp, TMPLD 109. The need for a pattern colours P.'s thought at all periods. Cf. 
Crat. 389 a, Euthyphro 6e. 
2 Phil. 28d-e, Tim. 29a, 39? (voOs). Conversely the ????? of Phil, is ?? ??????????? at 27b. 
voOv ... ????? ?????????? at Phil. 28 e, if the wording is pressed, might suggest that P. has not 
yet distinguished Demiurge from World-soul as he does in Tim. Cf. 30c-d and de Vogel, 
Philos. pt 1, 227; also Hackforth, CQ 1936, 7. 
3 It is interesting that Plato includes graphic artists (????????), for in one aspect the Demiurge 
is such an artist: the world is a copy B9 b), and 'the artist, clearly, can render only what his tool 
and his medium are capable of rendering. His technique restricts his freedom of choice. * (See 
Gombrich, Art and Illusion, 56.) 
254 
Maker, Model and Material 
Demiurge is sometimes described as 'making' or 'putting together', 
but also as 'ordering' or 'shaping' the elements which had hitherto 
tumbled in haphazard disorder.1 
Plato's first lesson, then, about the Maker of the world is that, as a 
demiourgos, he is not in sole and absolute control, but must bend to his 
will a material that is to some extent recalcitrant. Otherwise, being 
wholly good himself, he would have made a perfect world B9d-3oa). 
This is philosophy, not myth. Those who demythologize him away 
(and so to water down Plato's persistent theism seems to me quite 
unjustified2) are at least left with a universe whose fundamentally 
rational structure is infected by an irreducible element of imperfection 
and waywardness inherent in its bodily nature. But at this point modesty 
compels us to recall Plato's own words at 28c:' The Maker and Father 
of this Universe is hard to find, and when found, impossible to describe 
to all and sundry.' 
The model. Following, or preceding, the trend of the Philebus, the 
Forms are now (to use the familiar Aristotelian terms) formal causes 
only, having resigned to a separate power the quasi-efficient function 
which they, rather obscurely, possessed in the PhaedoZ This power, 
1 ?????? 3 ib, ?????????? 29 a and frequently, ?????????? 33 d, the cosmos a ???????? 48 a 
et ?/., ??????^????? 37c-d and 39? (with ????????????); but ?????????? and ?????????????? 
53b, ?????????? 69c. Cf. 69b ??????? ?????? ? ???? ... ?????????? ??????????. The function 
of the Cause is exactly the same as in Phil. Cf. ???????? ??? ??????????? at 30 c. The associations 
and history of the word kosmos have been discussed in vol. 1, esp. nof. and 208 n. 1. Reference 
should also be made to the following modern authorities: W. Kranz,' Kosmos als philosophischer 
BegrifTfnihgriechischer Zeit* in Philologus 1938-9, and 'Kosmos*, Arch. f. Begriffsgesch. 1958; 
H. Diller, 'Der vorphilosophische Gebrauch von ?????? und ???????*, Festschr. Snell, 1956; 
J. Kerschensteiner, Kosmos: quellenhit. Unters. iu den Vorsokratikern, 1962; A. Lesley, Kosmos, 
1963 (inaugural lecture); C. Haebler, 'Kosmos: eine etymol.-wortgesch. Untersuchung*, Arch.f. 
Begriffsgesch. 1967; J. Puhvel, 'The Origins of Greek Kosmos and Latin Mundus\ AJP 1976. 
* Cornford wrote {PC 34): 'Plato is introducing into philosophy for the first time the image 
of a creator god. * Did he not learn it from Socrates? According to Xenophon S. distinguished 
from the other gods ? ??? ???? ?????? ????????? ?? ??? ??????? {Mem. 4·3·*3)· He spoke of 
? ?? ????? ????? ????????? and made Aristodemus admit that the economy of the human body 
betrays the hand of a wise and beneficent demiourgos A.4.5 an<^ 7)· 
3 See pp. 213 f. above. The separateness of theah-?? is especially emphasized in Phil. It is always 
a ???????? besides the Unlimited, Limit and their mixture B3 d, 26 e, 3oa-b, a point against those 
who would identify the Demiurge with his model, pp. 259-62 below). In these dialogues there 
is a notable absence of the term ???????, a trouble-making metaphor even if P. only meant it as a 
variation for ??????? (?. 46 above), and one which took a severe beating in the Parm. Its absence 
is only confirmed by Grote*s attempt to deny it, for all he can quote is ????????????? ??? ?????? 
at 51a, which he says is equivalent (m, 268 n.). It does not even refer to physical objects but to 
255 
Timaeus and Critias 
being a demiourgos, must have worked to a model. Now at the very 
outset Timaeus reasserted, as the first essential to be grasped, the 
familiar1 Platonic distinction between what exists ungenerated and 
eternal, and what suffers coming-to-be and perishing but never fully 
exists; the one comprehended by the intellect through dialectic, the 
other only an object of belief (doxa) through sensation without  
reasoning.2 For Plato the question is: on which of these did the Maker fix his 
eye as model for the world? And one cannot claim that he makes much 
use of dialectic to settle it. 'If this world {kosmos) is fair and its 
demiourgos good, he clearly looked to the eternal: if otherwise, to what 
has become. But that would be a wicked utterance. It is obvious to all 
that he looked to the eternal, for the world is fairest of created things, 
and he the best of causes.' He need say no more at present. That reason 
alone lays hold of the perfect ungenerated Forms, and their function as 
paradigms of the whole sensible world, he has already taught in many 
dialogues; and that the world is a creation of intelligence is inferred, as 
in the Philebus (pp. 2i4f. above), from the regularity, beauty and 
evidence of purpose in the major cosmological events, night, day, the 
cycle of the seasons by which all earthly life is maintained, and the 
recurrent celestial movements on which they depend. This is expanded 
in the Laws (897 bfF.), the logos to Timaeus's mythos. The ' ideological 
argument', powerful through centuries of Christian apologetic down 
to the well-known hymn of Joseph Addison, stems from Plato no less 
than the psalmist. To both 'the heavens declare the glory of God, and 
the firmament showeth his handiwork'.3 Plato's argument is not that 
there is no ugliness or disorder in the world, but that they are local and 
insignificant compared with the marvellous organization of the cosmos as 
a whole. Without a rational God to tame it, disorder would have been 
the rule, not the exception. As it is, he thought, a philosophic observer 
the ???????, and the meaning is quite different. It may seem strange that Aristotle condemns the 
Forms so severely for not being efficient causes, when a separate efficient cause is so carefully 
provided. He actually asks: What is it that does the work with its eye on the Forms? (Metaph. 
991 a20-2). But to him a personal demiurge was just one of those 'empty poetic metaphors* of 
which in the same passage he complains. His own God does not soil his hands with craftsman's 
work: his mere existence is enough to sustain the ungenerated universe in being. 
1 Especially from Rep. 5 (vol. iv, 487 ff.). 
a 27d-28a. The distinction is so important that it is elaborately repeated at 5^3-5237. 
3 Also to Isaac Newton. See the impressive passage quoted by Cornford in Princ. Sap., p. 21, 
and in n. 2 to p. 286 below. 
256 
Maker, Model and Material 
cannot deny that the world is a product of intelligence and copies a 
perfect model as well as the limitations of physical embodiment allow. 
For Plato the endlessly repeated and (as he believed) perfectly  
circular motions of the cosmos and heavenly bodies were not only produced 
by, but actually resembled, the operations of Mind. (More of this later.) 
Sambursky in his Physical World of the Greeks (p. 54) has a comment on 
the difference here between the ancient attitude and our own. We live 
in a machine-age, and the essence of a machine is to reproduce the same 
movements exactly. Consequently we associate the idea of exact 
repetition with 'soulless mechanism'. But in an age of handicrafts, any 
exact reproduction appeared as a sign of the artist's divine inspiration. 
It sounds an illuminating comparison, but our attitude would not have 
surprised Plato, for it was current in his own time. In the Epinomis 
(982 c) we read that men ought to have regarded the uniformity of the 
stars' motions as proof of their divinity, but in fact ' most of us think 
the very opposite, that because they always do the same thing in the 
same way they have no life'. Could he question us now, he would, I 
think, ask: 'Do your machines, then, make themselves? Can they exist 
without a mind to design them?' Machines presuppose minds, however 
much the word 'mechanistic' may be misused to deny it. 
At any rate in Plato's view regularity in natural processes implied 
purpose. Arguing against the primacy of chance in some of the earlier 
philosophies of nature, Aristotle (no friend to mythI admits, strangely 
to our minds, only two alternatives: regular repetition demanding a 
ideological explanation, and chance events which are the exception. A 
hot day in the winter would be due to chance; hot days in the summer 
are produced by the normal natural methods; therefore nature is  
purposeful (Phys. 198b 34-9938). Here he agrees with Plato. 
The concept of the Model raises problems which are probably  
insoluble. God, says Timaeus, made the cosmos a living, thinking  
creature (p. 275 below), and his next question is: 'In the likeness of what 
living creature2 did he make it?' C0c). Not of any particular species. 
1 'The sophistry of myth is not worth serious consideration: only demonstration can instruct 
us* (Metaph. ioooai8). 
* Greek has the convenient word ????, from ien, to live, often rendered with perfect propriety 
'animal*. But some of the associations of animality in our language seem inappropriate here. 
? have adopted 'living creature*, but of course the Model is not a creature in the literal sense. 
18 
257 
GHG 
Timaeus and Critias 
It must embrace all intelligible living creatures, as the cosmos includes 
ourselves and all other visible animals. The living world is the image, 
or projection into body, of the Form of Animal and all the subordinate 
Forms of Man, Horse, Dog etc. Cornford has explained this well 
(PC40): The intelligible Living Creature Ms a generic Form containing 
within itself the Forms of all the subordinate species, members of which 
inhabit the visible world'. Its widest divisions are four C96-40a), sc. 
the visible gods (stars, planets, Earth), birds, fish, land-animals. 
These main types, as well as the individual species of living creatures and 
their specific differences, are all, in Platonic terms, 'parts' into which the 
generic Form of Living Creature can be divided by the dialectical procedure 
of Division. The generic Form must be conceived, not as a bare abstraction 
obtained by leaving out all the specific differences determining the subordinate 
species, but as a whole, richer in content than any of the parts it contains and 
embraces.1 It is an eternal and unchanging object of thought, not itself a 
living creature, any more than the Form of Man is a man. 
Cornford continues: 
Plato does not say . . . that this generic Form of Living Creature contains 
anything more than all the subordinate generic and specific Forms and 
differences that would appear in the complete definitions of all the species of 
living creatures existing in our world, including the created gods. We have 
no warrant for identifying it with the entire system of Forms. 
Yet the cosmos as a whole was made in the likeness of this supreme 
generic Form. Should it not therefore embrace the Forms of all that the 
cosmos contains?2 Thus Lee {Tim. 10): 'Plato must mean a complex 
system of Forms, containing within itself all the subordinate Forms 
whose likeness we can trace in the world of Becoming.' He finds the 
conception of the 'intelligible living creature', and its place in the world 
of Forms as a whole, not easy to grasp, and suggests that 'its presence 
is perhaps due as much to the requirements of the craftsman analogy as 
to any philosophic principle'. Plato, as Archer-Hind says, has left his 
1 The Platonic dogma that the higher (more universal) Forms were also the richer and fuller 
in content and being (elaborated on pp. 432 f. below) was one against which Aristotle reacted 
vehemently. For him it was axiomatic that only the individual was fully real (Cat. 2a 11-14), and 
the genus the comparatively unformed 'matter' or 'substratum' of the species. (Metaph. 103836, 
1058323 etc. See Bonitz, Index 125 a, 7873.) 
* Taylor's attempt to help by a reminder that jcoa in Greek could mean 'pictures', whether of 
animals or not (Comm. 81, adopted by Grube, PT 169), does not seem apposite in the context. 
258 
Maker, Model and Material 
intention uncertain, and the question is hardly to be decided now. At 
39 a Timaeus says that, wishing to make the world more like its model, 
God (or Mind, Nous) gave it the four kinds of living creatures,' seeing 
what and how many are the Forms1 in the Living Creature Itself. This 
supports the more limited conception of it. Even so, in describing it at 
30d he called it 'the finest and in all respects most complete of the 
intelligibles', which sounds more comprehensive.2 
Relation of the maker to his model. Here is another crux on which opinion 
has been divided down the centuries. Is their differentiation only 
mythical? Wilamowitz identified the Maker and Father with the Form 
of the Good in the Republic? Hager also, if I understand him, identifies 
God with his model, i.e. the Forms, but not with the Form of the Good, 
and de Vogel writes: 'He is, so to speak, the intelligible order turned 
towards creation and personified into a creating God and Father.' 
Archer-Hind, amid much un-Platonic language, asserted that 'all that 
exists is the self-moved differentiation of the one absolute thought, 
which is the same as the Idea of the Good', and 'the ?????????? 
[Model] is universal thought regarded as the supreme intelligible, the 
?????????? [Demiurge] represents the same regarded as the supreme 
intelligence'.4 The most frequent claim is that the Forms are nothing 
1 ?????. The word by itself could mean only species or kinds, but since they belong to the 
intelligible sphere they are of course Platonic Forms. 
* Cherniss however {AC? A 576) thinks Proclus right in supplying ^??? after ?????????. He, 
like Cornford, favours the view that the paradigmatic Living Creature contains only Forms of 
the animate, which Archer-Hind (Tim. 34f.) also thought more reasonable. On the other side 
one may mention Taylor, Comm. 80f. ('the complete system of the Forms') and de Vogel, 
Philos. pt 1, 181 ('l'ensemble des Ideas'). Another question which has troubled scholars is 
whether the Model, being a jcpov C0 c 3), is itself alive. To answer this dogmatically, it is ?? ? ???? 
jcpov C9 ? 8), i.e. a Form, through which living creatures have their ?????, and Forms do not 
have ?????. For the sense in which Forms are self-predicable see pp. 43k and 47 above, 
and for arguments favouring the contrary of what I have said, de Vogel, Philos. pt 1, 228 f. 
3 ? ??? ?????? ???? or ?? ?????? (Rep. 508 b and c etc.), so prominent in the Rep., is not in fact 
mentioned in Tim. 
4 Wilam. I, 605; F.-P. Hager, Der Geist u. das Eine 37-43; de Vogel, Philos. pt 1, 229; A.-H. 
Tim. 28, 95 n. How the intelligible order turned towards creation de Vogel does not explain. 
If the Forms made themselves sole efficient causes, P. would be more naively susceptible to 
Aristotle's criticism than he was. 
Taylor saw in A.-H. a 'determination to force on Plato a philosophy of his own devising' 
(Comm. 38). It was a philosophy very much ot his own time, but this conception of the two aspects 
of thought has persisted. Hager in 1969 (o.c. 39) wrote of 'das wahrhaft erkennbare, bzw. 
wahrhaft erkennende Sein*. (Cf. p. 40, where the puzzling phrase 'rein theoretisch oder der 
Moglichkeit nach' does little to modify his statement.) 
259 
18-2 
Timaeus and Critias 
but thoughts in the mind of God, who creates them by thinking of 
them. So said Philo of Alexandria and many Christian Platonists, 
notably St Augustine, and it has been affirmed by good Platonic 
scholars of the nineteenth and early twentieth centuries such as Ritter, 
Henry Jackson and Archer-Hind, and as recently as 1969 by Moreau. 
Yet the case against it, as put for instance by Audrey Rich in 1954 (and 
for that matter Martin in 1841), has never really been answered. Others 
who have denied it include Cornford (' The model, as strictly eternal, is 
independent of the Demiurge'); Skemp ('The ?????????? is not to be 
confused with the outojcoov [Form of Living Creature], which is the 
object of his contemplation'); and Taylor, who also refused to regard 
the distinction as 'only part of the fanciful imagery of the dialogue'.1 
In Rep. 6 it is said that the Form of the Good gives both existence 
and essence2 to the other Forms, and this is often compared to the 
activity of the Demiurge in the Timaeus as only another expression of 
the same thing. But the two accounts have nothing in common. Rep. 6 
says nothing of a planning Mind, and has no concern with the creation 
of a cosmos. There is no hint, there or elsewhere, that the Form of the 
Good, or any Form, is or has nous, which is the whole being of the 
creative Cause of the Timaeus. The centre of Plato's interest has shifted, 
and his metaphysical scheme is now that of the Philebus.l Another 
passage on which some have relied is Tim. 29 a, where it is said that the 
Maker wanted all things to be like himself. But, runs the argument, he 
made it like the Model; ergo he and his Model are mythical expressions 
for the same thing.4 in context, this stretches the words much too far. 
1 For the earlier scholars see Rich's neglected article 'The Platonic Ideas as the thoughts of 
God', Mnemos. 1954, 123 n. 1. Other reff. are to Moreau, 'The Platonic Idea', IPQ 1969, 509^, 
which I find hard to reconcile with 511 n. 135, where he speaks of the Ideas within the archetypal 
Living Being as 'the very object of the Intellect', and of the intelligible order as perceived by God 
and become 'the rule (????) of his actions' (my italics); Cornford, PC 40f.; Skemp, TMPLD 
108, and cf. 115; Taylor, Comm. 81 f. Add Brochard, Etudes 95-7. Brochard criticizes Lutoslawski 
on the point in his Etudes, 166 n. 1. 
2 ?? ????? ?? ??? ??? ??????, Rep. 509b. 
3 Cf. Thompson on Phil. i6c(JPh 1882, 20): 'Those whom the magnificent language applied 
in the Republic to the '?????? may have tempted to believe that the God of Plato was, if not a 
number, an Idea, will find I think a corrective to that misapprehension in the passage quoted from 
the Philebus (p. 26E seq.)· The language is indeed so explicit as to seem designed for the purpose 
of obviating the very inference I have alluded to.' 
4 E.g. Hager's argument is (o.c. 43):' Die Welt nach Platon nicht zwei verschiedenen Wesen- 
heiten ahnlich sein und werden kann.' 
260 
Maker, Model and Material 
The Maker, says Timaeus B9a~3ob), being good, and so incapable of 
jealousy, wanted everything to be as like himself as possible. This is the 
true reason for the genesis of the world. Wishing everything to be good, 
and nothing bad, so far as might be, he took over matter with its restless 
and discordant motion and reduced it to order. Judging, moreover, that 
anything of a visible nature was better with reason than without, he 
made the cosmos a living, rational creature. That is all. The world 
resembles its Maker simply in being (a) good and (b) alive and rational. 
To achieve this end he modelled it on the eternal Forms B9 a). 
In describing the Forms as thoughts of God, some speak in terms of 
the Aristotelian concept of thought which thinks itself. So Archer- 
Hind: 'Thus does dualism vanish in the final identification of thought 
and its object... Thought must think: and since Thought alone exists, 
it can but think itself.'1 Aristotle's supreme God, unlike Plato's, is 
wrapped in eternal self-contemplation to the exclusion of any providence 
or planning of a world, on the argument that the perfect being's thought 
can only be of what is best, and that is himself. How he does this is 
explained on the lines of Aristotle's general psychology. In sensation 
and thought the psyche assimilates the form (sensible or intelligible as 
the case may be) without its matter. (Form is of course internal for 
Aristotle.) Both are purely psychical functions, though in sensation the 
psyche employs the bodily organs as instruments. In sight the eye, a 
physical organ, becomes coloured, but the psyche becomes aware of 
colour. In thought the psyche assimilates directly the intelligible form, 
that is, the definable essence of the object. But taking on a form is not 
an indifferent act like putting on a coat. As the actualization of a 
potency it implies a change in the object informed. In thinking of  
something, the mind becomes that object in so far as it is an object of thought, 
i.e. assumes its intelligible form. That is why, as we think of things or 
people, we usually suppose thought to be of what is outside the mind, 
not identifying thought and its object completely, because there  
remains the material element which of course the mind does not absorb.2 
Consider however the special case of God. As perfect being, he is pure 
1 A.-H. o.c. 28. 'since Thought alone exists' is surely a perfect example of petitioprincipiiy the 
assumption of what has to be proved. 
2 This applies even when we think about ourselves: ????? ?* ?? ???????, Metaph. 1074^36. 
261 
Timaeus and Critias 
actuality, with no unrealized potentiality at all. Hence the object of his 
thought (himself) has no matter,1 and therefore only intelligible form. 
If the mind in human processes of thought becomes identified with the 
intelligible form of its object, the identification of mind and its object 
must in this case be complete. Add the fact that the act of thought is 
eternal, and there is no longer any distinction between thought and its 
object: their essence is one and indivisible.2 
This being so, there is much to be said for Audrey Rich's suggestion 
in the aforementioned article (p. 260 n. 1) that the notion of the Forms 
as thoughts of God in Plato originated in the desire of later antiquity to 
reconcile the Platonic theory of independent Forms with the Aristotelian 
doctrine of immanent form.3 Our excursion into Aristotle's psychology 
may have seemed out of place, and certainly its brevity does Aristotle 
scant justice; but I hope it has been clear, and it was needed to  
emphasize by contrast how impossible it would have been for Plato, with his 
different metaphysical and psychological assumptions, to reduce the 
Forms to thoughts in the mind of the Creator. In every dialogue in 
which they appear, their existence independent of any mind conceiving 
them is a leading feature. The craftsman analogy prevails; and whereas 
for Aristotle the form of his product must pre-exist in the craftsman s 
mind* for Plato he looks (at two removes perhaps) to an external, fixed, 
objective standard or Form, which he attempts to reproduce, though 
with imperfect success. The last three words apply even to the divine 
Craftsman who made the world, for he too had to work with a material 
which he could only partially subdue. To this we must now turn. 
The material D8d~53c). Many have bewailed the obscurity of much of 
the Timaeus. In this case the warning—a strong one—comes from 
Plato himself. At 486-49a Timaeus says he must make a new start. 
Earlier it seemed as if it was enough to speak of a model and its copy, 
but now the logos demands something else, a * dim and difficult concep- 
1 Matter is potentiality, and form actuality (De an. 41239). 
2 For God's activity as thought of himself see Metaph. ?, chh. 7 and 9. Ch. 9 alludes to the 
general doctrine of the nature of thought which is expounded in De an. For sensation see bk. 2 
ch. 12 and for thought 3.4, esp. 43032-9. 
3 R. M. Jones has also written on the origin of the notion, in CP 1926. 
4 Metaph. 1032b ? ???? ?????? 5? ???????? ???? ?? ????? h ?? ????. 
202 
Maker, Model and Material 
tion\ Besides these two there must be that in which becoming takes 
place. To explain it is hard. After a first attempt he must * try to put it 
still more clearly' E0 a), but at 50 c copies of the Forms enter into it 'in 
a strange manner hard to express'. Again it is 'invisible and shapeless, 
receiving all things, partaking in some most bewildering way of the 
intelligible and hard to capture' Eia-b); 'indestructible . . . touched 
without the senses by a sort of bastard reasoning, hardly credible' 
E2a-b). In the effort to portray it Plato resorts to simile after simile, 
and it will be prudent to marshal the various expressions used as he 
struggles to convey this mysterious notion from his own mind to ours.1 
1. It is an ultimate, having existed, like Being and becoming, 'even 
before the world was born \2 
2. Its nature and function are 'to be the Receptacle and so to speak 
nurse of all becoming' D9b). There being no obvious similarity  
between a receptacle and a nurse, Plato might have done better to omit 
'nurse' and keep to the mother-image. ('Nurse' recurs at 52d5.) 
3. It is a plastic substance3 capable of receiving impressions, moved 
and shaped by what enters into it E0c). 
4. The previous image is both extended and illustrated. A soft 
material which is to have a device impressed on it must first be made as 
smooth as possible Eoe). At 5oa-b the illustration is gold, which a 
goldsmith fashions into many different shapes. If asked of any of them 
'What is that?', it would be safest to reply 'gold', not 'a ring' or 'a 
triangle', for even as the words were spoken, he might be giving it a 
different shape. So with 'the nature that receives all bodies'. It is only 
itself, not any one of them. If it resembled any of the things it takes in, it 
would reproduce other things badly, intruding its own features as well. 
1 It has been an object of controversy down the centuries. Baeumker in his book Das Problem 
der Materie B. Abschnitt, 1 ioff.) includes an extensive review of opinions from Aristotle through 
Plutarch and the Neoplatonists down to his own time. 
2 52d. There was of course ??????? before the ??????* of our world: the act of the Demiurge 
was simply to introduce order into it. The question of the temporal origin of the world will come 
up later (pp. 299-305). For the present those who like may take * before* as referring only to 
logical priority. 
3 ????????? E0C2) is used at Tht. 191c of a smooth block of wax before impressions are 
stamped on it, but at i94d of impressions already stamped. This second sense, of * moulds* or 
*dies*, recurs at Laws 800b, where certain typical cases are to be used as ???????? for legislators, 
and is probably the commoner. Here however only the first sense is intended. The receptacle 
must be ??????? ?????? ??? ????? ???? ?????? ???????? Eod). Cf. the use of ??????????? at 
5od. 
263 
Timaeus and Critias 
5. The same point is made by comparing it to the oils used as a base 
for perfumes. They must themselves be as far as possible scentless 
Eoe5-8). 
6. 'The Recipient is fittingly likened to a mother, the model of 
becoming to a father,1 and the nature that arises between them to  
offspring. ' It was a common Greek belief that the father was the sole 
cause of generation, the mother contributing only nourishment and a 
place for the embryo to grow in.2 Hence 'mother and receptacle' at 51 a 
go quite naturally together. There is also a distinct trace of the 
Pythagoreans, with their association of the unlimited with the female, 
and the unit, the principle of limit and order, with sperma (Arist. 
Metaph. 1091a 16; vol. I, 245 f., 278). 
7. After all the foregoing, Plato refers to the Receptacle as 'space' 
(????), and says that it provides a 'seat' (????) for everything that 
becomes E2a8-bi). It is what we have regard to when we imagine^ 
that all that exists must be in a certain place and occupy space E2a-!)); 
and in summing up at 52d he refers to his three factors as Being, Space 
and Becoming. 
8. It is in constant irregular motion, swaying and shaken 'like grain 
in a winnowing basket' by the 'powers' of the rudimentary elements or 
qualities that pervade it, and in turn reacting on them E263-5). 
What are we to make of this dim and dubious something, scarcely an 
object of belief, let alone of knowledge? Its essence lies in the oft- 
repeated statement, backed by the similes of the wax for stamping and 
the base for perfumes, that in itself it must have no perceptible qualities. 
Because its function is to 'receive' sensible copies of the eternal realities, 
it must be without any form4 of its own which would distort the image. 
It is the medium or material 'in which' (again a repeated phrase) all 
1 5od. ?? ???? refers back to the previous sentence, and is, more literally, 'that whence  
anything that becomes has its development by being made like to it*. Surprisingly, those who 
maintain the identity of the Demiurge with his Model have not, so far as I know, availed  
themselves of this comparison of the latter to a father. 
a Cornford illustrates from Greek literature, PC 187. Aristotle calls the menstrual fluid the 
'matter' (???) of the child, whereas the semen provides the efficient cause and the form (GA 
?-?)*-*,·!, 738b2o). 
3 ??????????????. It is a 'dream* because of course what really exists (i.e. Forms) is not 
in space. For this figurative sense of dreaming cf. Rep. 476c, 534c. See also Baeumker, P.d.M. 
139, and for the hyperbaton Cornford, PC 192 n. 2. 
4 ?????, ????? and ???? are all used in this non-technical sense E0c 1, d7, 5ia2, 7). 
264 
Maker, Model and Material 
perceptible things or qualities come to be. 'Receive' (????????) here 
has a wide sense, especially that of taking an impress or assuming a 
character. Plato had not developed Aristotle's technical vocabulary of 
potentiality and act, but it is not misleading to say that the material has 
in and by itself (in which naked state it never occurs, any more than 
Aristotle's prime matter, of which, as in Plato, the simplest kinds of 
body consist1) the potentiality of being informed by the properties of 
fire, air, water or earth.2 The phrase ' in which' certainly does not, as 
Baeumker claimed, confirm his view that the receptacle is merely 
'empty space, sheer extension', for it is ambiguous in Greek as in 
English. A clay bust is in space, but it is also modelled in clay.3 
The third postulate, then, is something that can be called a matrix 
(?????????) or alternatively space (????).4 Many good scholars, from 
Zeller onwards have insisted that Plato intended it to represent nothing 
but empty space, or extension, and some have compared it to the matter 
of Descartes.5 'That the so-called "primary matter" of the Timaeus is 
1 It is becoming fashionable to deny that Aristotle believed in a 'prime matter*. See e.g. 
Charlton's App. to his ed. of Phys. ? and 2, pp. 129-45 (criticized by Owens in Phoenix 1971, 
281 f.). But I continue to regard as such the ??? . . . ?? ?$ ??????? ?? ????????? ????????, also 
called ?? ??????? ???? ???????? (GC 329326 and 33)· ?. R. King's article in J HI 1956, 370-90, 
I find quite unconvincing; but these are matters for the next volume. 
2 Baeumker gives no grounds for his assertion that 'Der Begriffdes bloss moglichen seins, auf 
den Aristoteles das Wesen der Materie zuriickfuhrt, ist dem Plato noch fremd' (P.J.M. 186). 
P. has no technical term for it, and confesses that he is trying to express a novel and difficult 
conception. In a different connexion he clearly conveys the distinction between potential and 
actual knowledge at Tht. 197b-d. (Cf. Arist. EN 1146b31 and Taylor, PMW 343.) 
3 Baeumker, P.d.M. 166. He was following Zeller A1.1, 734 n. 1) and has been followed by 
Solmsen (ASPIV 122) and Cornford, who writes (PC 181): 'There is no justification for calling 
the Receptacle "matter"—a term not used by Plato. The Receptacle is not that "out of which" 
(?? ??) things are made; it is that "in which" (?? ?) qualities appear, as fleeting images are seen 
in a mirror.' But cf. 50? 2? ????? ??? ??????? ??????? ???????????, whereas ?? is used of gold 
to convey the same notion. (The false distinction is again drawn by During, Aristoteles, 31?. 202.) 
The indifferent use of the two expressions in connexion with raw material is exemplified by their 
appearance together at Phil. 59 ? and Pol. 288 d. Keyt too (AJ ? 1961, 298) thinks that comparison 
with a mirror 'best captures Plato's thought'. But it is noteworthy that although P. uses a rich 
variety of metaphors to convey his difficult conception, and although (as K. notes) the mirror- 
metaphor is one of his favourites, yet 'The third factor is never, in fact, called a mirror' (Keyt, 
l.c). 
4 These are the only two words applied to it without qualification or hint of simile (such as 
???????, ????, ????????^???). I would not mention something that may well be accidental, were 
it not that Baeumker (o.c. 184) claims it as conferring special status on ???? without mentioning 
?????????. 
5 E.g. Baeumker, o.c. 187; Taylor, Comm. 312, 322, 387; Milhaud and Robin (Claghorn, 
ACPT 15 n. 19). But Brochard wrote (?tudes 108): 'Quant a voir dans Platon un pr?curseur de 
Descartes, ce n'est pas possible.' His reason is that there is nothing geometrical about Plato's 
265 
Timaeus and Critias 
space in three dimensions and nothing else', declared Burnet (T. to P. 
344), 'is really quite certain both from Plato's own language and from 
the statements of Aristotle.' But it is time to heed the advice of Timaeus 
D9a-b), that to understand the Receptacle we first need a 'firm and 
trustworthy account' of fire, air, water and earth, which Plato, like 
Empedocles, regards as the primary bodily constituents of everything in 
the physical world. No one, he says, has explained their origin, and 
we talk of them as ultimates without knowing what they are. Far 
from being stoicheia? they are even more complex than syllables. 
What, we must ask, were they like before the creation of the cosmos? 
D8b-c.) 
First, observation shows that they are unstable, constantly changing 
into one another by condensation and compaction and rarefaction or 
dispersion—the old Ionian view. We cannot therefore call any of them 
an existing thing, but only a temporary qualification, a 'such', not a 
'this'.2 But qualities presuppose something in which and from which 
they appear and disappear.3 Plato, one may say, is struggling to express 
Receptacle. This is certainly right, and Taylor wrong in calling it 'geometrical extension'. Popper 
gives a helpful summary of the Cartesian theory in his article in Studies in the Philosophy of 
Biology 1974, 262. Note (especially in view of what will come later) that he does not speak of 
Descartes reducing matter to pure extension but to 'extended substance'. Matter was not reduced 
to space, but 'space too was reduced to matter, since there was no empty space but only the essential 
spatial extension of matter'. (My italics.) Popper sees Plato's Receptacle as similar, a 'liquid-like 
medium' in which vortical motion could take place without empty space (C. andR. 81 n. 22). 
1 Elements or letters, pp. 176, 211 above. Unlike Empedocles, the atomists did try to penetrate 
beyond the four elements, but in Plato's view gave the wrong answers. Serious students of 77m. 
should be warned that this passage (roughly from 49b to 50b) has been the subject of prolonged 
controversy. Cornford's interpretation in PC A78-80) was challenged by Cherniss in AJP 1954. 
Gulley in AJP i960 found Cherniss's view 'self-refuting and incorrect' and in the same journal 
for i960 ?. ?. Lee had something to say about both Gulley and Cherniss. 
* Owen's objection (SPM 323) that even the word ???????? cannot consistently be used, and 
contradicts Crat. 439d 8-9, has I hope been dealt with on pp. 79-82. (Cf. also next page, on 77m. 
52a 5.) We must not confuse the Heraclitean flux of a world with no unchanging Forms behind 
it ('if even Beauty itself gives us the slip', Crat. l.c.) and the same flux in a scheme (Plato's)  
including the Forms. Since this chapter was written, a new interpretation of the relevant passage and its 
context has been offered by D. J. Zeyl in HSCP 1975, with detailed criticism of other recent views. 
3 I have avoided the word 'substratum', just as P. lacks the Aristotelian ????????????, but is 
not this what we usually mean by the word? Taylor (Comm. 387) forbids us to 'introduce from 
Aristotle the notion of "matter" as a substratum of events. Aristotle is quite explicit . . . that 
Timaeus knows of no "matter" distinct from ????.' He gives no reference, but may be thinking 
of Phys. 209b 11, where A. says ??? ??? ? ??? ??? ????? ?????? ???? ????? ?? ?? ??????. This 
means that (as is obvious) he identified matter and space, not that he replaced matter by space. 
It is an interesting coincidence of vocabulary that matter (???) is for A. 'in the strictest sense' the 
substratum which receives becoming and perishing (?? ???????????? ??? ???????? ??? ?????? 
????????, GC 320a2). The idea that Aristotle's equation of P.'s ????????? with his own ??? 
266 
Maker, Model and Material 
for the first time what Kant said in 1787 and William Whewell sixty 
years later. 
If we remove from our empirical concept of a body, one by one, every feature 
in it which is [merely] empirical, the colour, the hardness or softness, the 
weight, even the impenetrability, there still remains the space which the body 
(now entirely vanished) occupied, and this cannot be removed. Again, if we 
remove from our empirical concept of any object, corporeal or incorporeal, 
all properties which experience has taught us, we yet cannot take away that 
property through which the object is thought as substance or as inhering 
in a substance.1 
There follows immediately in Plato the simile of the gold wrought into 
different shapes. Like it, the 'recipient of all bodies', 'of everything 
generated and perceptible' E0b, 51a), remains itself, uncommitted to 
any of the properties which enter into it, and which Plato now reveals 
as copies of the eternal realities2 Eob-c); for there is an intelligible 
Form for each of the primary bodies. Fire and the rest in the physical 
world take the names of their Forms and, though generated, in constant 
motion, appearing and vanishing in a particular locality, resemble these 
changeless realities.3 Here we have the familiar relationship of imitation 
in the classical doctrine of Forms, which Plato never abandoned, but 
has refined in three ways: 
A) The dubious notion of 'sharing' is dropped. 
B) The anonymous questioner of the Phaedo is given a final answer. 
Opposite qualities cannot change into one another, nor is it exact to say, 
rests on a misunderstanding is still strongly held. Cf. Solmsen in Mus. Helv. 1976, 27, citing 
Cherniss, ACPA i6^ff. 
1 Kant, Crit. of Pure Reason, Introd. to 2nd ed., § 2, trans. Kemp Smith. (The passage will be 
found in Edwards and Pap, Mod. Introd. to Phil. 3rd ed., 688.) Cf. Whewell, Phil, of the Ind. 
Sciences, 2nd ed. 1847 (new impr. 1967), 404 f. The notion of substance as a substratum of change 
is unpopular with most modern scientists as it was with Berkeley and Hume, and Dingle in an 
article in ? J PS 1951 dismisses it as pre-scientific and childish. Von Weizsacker however speaks 
more cautiously in W.-V. of Phys. 31-3, and in America E. J. Nelson undertook in 1949 an 
impressive defence of substance in this sense as indispensable to empirical knowledge. (See 
Philos. for the Future, ed. Sellars etc., 106-24.) Stebbing wrote of the permanent psychological 
need for such a conception in MIL, 404. Dr G. E. R. Lloyd has suggested to me that there is a 
distinction to be drawn here between the application of the idea of substance to ordinary  
perceptible objects (where without some such idea it would be difficult to make sense of our  
experience) and the physicists' researches into matter and energy, which have largely outgrown 
it. Dingle and Nelson could both be right. 
2 ???????? 50C5, ??????????? 51 a2. 
3 ???????? ?????? ?? 5*a5· 
267 
^imaeus 
as was said there, that 'things' (????????) take on these qualities in 
turn.1 A neutral substratum is called for, which by receiving the  
imprint of the Forms produces visible and tangible things or bodies. 
C) Interpreters of the Phaedo have differed over the status of 'the 
tallness in us' at i02d, and the sense in which, in this and other central 
dialogues, the Form was 'present' in the particulars, about which Plato 
was deliberately vague (iood). Some have taken the immanent tallness 
to be not the Form itself, but something lower in the ontological scale.2 
Timaeus is unambiguous. The Form 'neither admits anything into 
itself nor enters into anything else' E2a); it is only images or copies of 
the Forms3 that enter and leave the Receptacle, making it part fiery, 
part wet and so on. 
Some of those who interpreted the Receptacle as 'matter' destroyed 
their case by equating matter with body (Korper\ thereby ensuring that 
subsequent criticism would be misdirected. Of course it is not body 
(which has sensible properties) any more than Aristotle's' prime matter' 
is body. It is an abstraction,4 reached by analysing corporeal substance 
in a way deemed necessary by both philosophers to escape Parmenides 
and explain the fact of change. It suggested to Plato the concept of 
space, not empty space as some have thought,* but ever full of a  
primitive kind of bodies, moving in every sense of the word  
kinesis—changing, generated and perishing, toppling over each other in their lack of 
homogeneity and balance E2 c) and communicating this motion to the 
Receptacle itself. That the Receptacle itself should be agitated, and 
communicate its agitation back to its contents, is the strongest  
argument for supposing that Plato meant what he said when he called it not 
only space but a matrix, 'stuff without property', as Popper describes 
1 Pho. io3a-c; vol. iv, 356. 
* Thus Ross wrote (PTI30):' What is present in the particular thing is not, strictly speaking, 
the Idea, but an imperfect copy of the Idea', and Rist and Cornford have taken similar views. 
(See p. 48 with ?. ? above, and Cornford, P. and P. 78.) I myself see strong reasons against this, 
but have felt a doubt whether Plato's own mind was clear on the point (p. 41). 
3 ??? ????? ??? ???????? 50c 5. 
4 I.e. something which we abstract or separate in thought from that from which it is inseparable 
in fact, not a mere thought or product of the imagination, 'bloss einen vorgestellten oder logischen 
Raum* (Gauss, Handk. 111.2, 198), nor yet an abstraction in Cornford's sense {PC 203). 
5 Especially in Germany: 4das Leere' or 'leerer Raum', Zeller 11.1, 740, supported by Baeum- 
ker and others. But there is no void in the cosmos E837, 79b 1, 80C3 and p. 290 n. 3 below), 
nor, I think we may assume, in the chaotic mixture which preceded it. 
268 
Maker, Model and Material 
it, comparing Anaximander's Apeiron (OS I, 211). This chaos, disposed 
'without reason1 or measure', is what the Demiurge took over: 'fire 
and water, earth and air, showing traces2 of themselves but in such a 
condition as might be expected of anything from which God is absent' 
E 3 b 3-4). Of their container, so far as its nature can be grasped, one 
can only say that the ignited part of it appears as fire, the wet part as 
water and so on E1b); so it offers every appearance to sight E2c 1), 
though considered in abstraction from the diverse properties that 
everlastingly pervade it, it is of course completely imperceptible.3 
What exactly 'enters and leaves' the Receptacle?* 
'No Form enters into anything else.' This accords with the purely 
paradigmatic role of the Forms in these later dialogues, but it is difficult 
to understand Plato's exact intentions. Has he really explained the 
relation between Forms and phenomena? What are the 'copies' which, 
by entering the Receptacle, impregnate it with fieriness, wateriness etc.? 
'Not the qualities, but enmattered forms', said Proclus oracularly,* but 
this hardly sounds like Plato. With the Forms as patterns only, he 
premises both in the Philebus and here a separate efficient Cause to 
create a cosmos in their image. This is Mind or God, but so far there is 
no cosmos and God has not yet taken the chaos in hand. His confession 
that the copies of the Forms ' take their stamp from them in a mysterious 
and scarcely explicable manner' E0c) betrays his embarrassment as a 
philosopher. He has not lost faith in the transcendence of the Forms, 
a partly religious belief as above all the Phaedrus showed, but one may 
1 Or proportion, ??????. Cf. our use of'irrational' to mean 'without ratio* (Popper C. andR. 
84). 
2 ????, primarily footsteps, and so traces of what is past rather than, as here, the beginnings 
of future development. (But cf. the use at Pol. 301c) But Plato with his figure of the ????????? 
in mind thinks of them as imprints on the Receptacle. With ???????? and ?????????? cf. 
Theocr. 17.122 ????? ?????????? ???? (of dust). Of this inchoate state P. says at 69b that there 
was nothing then deserving the names we now apply—fire, water and the rest. 
3 This should make it clear that there is no contradiction, as some have claimed, between 
'invisible', 'grasped without sensation' E137, 52b2), and on the other hand, 'appearing in 
every guise to sight' E2??), together with the statement that God took over 'all that was 
visible' C032). 
4 ?? ???????? ??? ???????. It is terribly difficult to avoid the word 'things', and once again 
one envies P. the Greek resource of using the neuter plural participle only. 
5 §???? ????. See Cornford, PC 183. 'Enmattered forms' are an Aristotelian conception, 
though not as Cornford says an Aristotelian phrase, ????? 2????? occurs once, at De an. 403325, 
of the ???? ?????. 
209 
Timaeus and Critias 
be excused for feeling that the time is ripe for Aristotle to come forward 
with his conception of form as the intelligible, definable element in 
things. It helps little to suppose that the pre-cosmic chaos never actually 
existed, but only depicts what the world would be like if it were not 
divinely ordered. The analysis leaves us with two categories, both 
resembling Forms, of which only one seems wanted: the copies 
(????????) whose presence in the Receptacle of Becoming gives it 
visible and tangible character, and the physical bodies, or ' things which 
become', also Mike' the Forms and named after them, compounded of 
the copies and Space together.1 No doubt the copies would be in 
modern terms what Crombie calls them (EPD ?, 303-5), property- 
instances as opposed to properties as such; but to Plato the Forms were 
never just properties, and I doubt if his mind was altogether clear on 
this point. Deprived of the almost mystical glow of conviction and the 
religious language that goes with it in the great central dialogues, and 
transferred to the ambience of an Anaximandrian Apeiron^ the relation 
between the Forms and the natural world becomes difficult to explain. 
But Forms remain for Plato the only possible bridge between a 
Heraclitean lack of all stability and the immobile unity of Parmenides. 
Either of these by itself would make knowledge impossible and cannot 
therefore be entertained by a philosopher. 
1 Is this pressing analogies too far? Archer-Hind says simply: 'The sensible objects of  
perception are the ???? ???????? ??? ???????' G7m. 45» mv italics). This may be right, but P. seems 
to say that, then as now, perceptible objects are bodies, albeit in an even more fluid state, whereas 
the ???????? ??? ??????? are only the formal characteristics of earth, air and the rest (?????? 
52d6). 
The ontology and cosmology of Phil, and 77m. undoubtedly shed light on each other, but— 
at least concerning the primitive chaos—I think it is a mistake to expect a one-to-one  
correspondence between the features of each. Archer-Hind's is a good attempt, but for one thing it is 
obvious from Phil. 23 c 9 and 2432 that no difference is intended between ????? and ?? ????? 
§???. Further, the temporal aspect of creation (whether literal or mythical) colours Timaeus's 
account and forbids exact equivalence. If the element of ????? is represented in 77m. by the 
infiltrating ???????? of the Forms, we have the problem that the marks of ?*??? include measure, 
proportion and number. These are expressly excluded from the primitive chaos, to be added by 
God E3a-b), whether we think of it in P.'s terms as existing 'before the genesis of the world* 
or simply as an imaginative description of the cosmos minus its rational organization. For this 
reason Ross's idea that the ???????? ??? ??????? are geometrical figures is impossible (Arist. 
Metaph. vol. I, 168). 
27O 
Cause of pre-cosmic motion 
What is the cause of the pre-cosmic motion? 
Here is another question on which scholars are, and will remain, 
divided.1 Timaeus's reason for the motion in the Receptacle of  
Becoming is purely mechanical E2e): 'Because it was filled with powers2 
neither alike nor evenly balanced, no part of it was in equipoise, but it 
was everywhere swung and shaken unevenly by them, and by its 
movement shook them in turn. '3 Plutarch however attributed it to soul, 
because 'soul is the cause and origin of motion' (De an.procr. 1015 c): 
the soul of the world was at first irrational, and until it was endowed 
with reason by the Demiurge its motions were disorderly. So Cornford 
(PC 205):' Since no bodily changes can occur without the self-motions 
of soul, the other factor present in this chaos must be irrational motions 
of the World-Soul, considered in abstraction from the ordered  
revolutions of Reason.' But even a myth (if this is all mythical) should be 
internally consistent, and in Timaeus's story the disorderly motion was 
there before the world-soul was created.4 Plutarch, like others, had in 
mind the Phaedrus B46 c) and Laws (896 a-b), where Plato says that 
soul is the cause of all motion whatsoever in the present world-order. 
With due deferences I would suggest that one should not try to press 
into literal agreement the words of three very different dialogues, in two 
of which at least Plato's fertile imagination is conveying his message 
through different pictures. Only in the Timaeus does he speak of a 
period before the cosmos was organized, and in both the others he is 
describing the nature of our world as if it had existed everlastingly. In 
the Laws (896e~97b) the motions of soul are all properly psychical, 
including wish, reflection, care, counsel, judgement true and false, and 
emotions. It may exhibit reason or folly, guiding things well or badly. 
In either case, with these and similar motions as primary, it 'takes over' 
1 To name only two examples, the view taken here agrees with Crombie's (EPD n, 227 f.) 
rather than Skemp's (TMPLD 76, in n. 1). 
2 I.e. qualities, hot, cold, wet, dry etc. (vol. 1, 325 n. 1). At 5033 P. uses the old word 
'opposites*. 
3 57? ?????? ??? ?? ????????? ??????? ?? els ??????????? ??? ??????? seems to clinch the 
point. Cf. Spoerri, R, de PhiloL 1957, 213. 
4 The above was written before J. S. Clegg's article 'Plato's Vision of Chaos* appeared in 
CQ 1976, but having read it, I see no reason to modify what I have said. 
5 Taran has presented a detailed case for the contrary view in § ? of his article in Anton and 
Kustas, Essays. 
271 
Timaeus and Critias 
the secondary motions of bodies and brings about their growth and 
decay, temperature, textures, colours and flavours. But (his favourite 
plea) only a rational soul could produce the regularity of the celestial 
motions (cf. 967b). No great ingenuity is needed to translate this into 
the terms of a genetic account in which, just as there was a pre-cosmic 
period, so there was a period before physical motions were due to the 
wish or judgement of any soul, good or bad. Certainly the Timaeus 
contains no hint of any other doctrine. They just 'came about of 
necessity* D7c), whose 'nature it is to cause motion' D76-48 a).1 
Finally, that motion in the Receptacle should be due to inanimate 
necessity accords with the scheme of Democritus on which Plato based 
it (p. 274 below). 
Plato's point is the same throughout, that our world is the product 
of reason and design, not chance. Here his target may be Empedocles, or 
possibly, as elsewhere, Democritus. It was Empedocles who attributed 
motion specifically to overbalancing.2 But whereas Empedocles was 
describing the origin of day and night, which he attributed, like all 
phenomena of the present world, to chance,3 for Plato no soulless  
process could produce such uniform recurrence. The sum of things must 
remain in turmoil and disorder until Mind intervenes to rescue it. 
Necessity 
We have seen that the Demiurge is not omnipotent, but must create his 
cosmos in a given material which cannot achieve the changeless  
perfection of the intelligible worlds This resistance to perfect ordering Plato 
ascribes to necessity {ananke), which he presents under two aspects, 
positive and negative. To take an example, God wished us to have sight 
so that by observation of the heavens we might be led to philosophy 
D6e-47c). That is its primary cause. But it was only possible through 
1 It is ? ??? ?????? ??? ???? ?????? ??? ?? ??? ?????? of Phil. 28 d. 
2 Emped. DK a 30 (vol. 11,186) ??? ????????? ??????????? ??? ?????. According to Aristotle 
(fr. 208 Rose, DK 68 a 37) Democritus used the more general term ????????. Cf. 77m. 52? 
???' ??????. 
3 Frr. 59.2, 103, 104 etc. See vol. 11, 161-4. So of course did most of the earlier cosmologists 
except Diogenes of Apollonia (vol. 11, 369), and above all Democritus. 
4 A point made in the Pol. B69 d): 'What we call universe (???????) and cosmos has received 
many blessed gifts from its progenitor, but nevertheless it partakes of body, and so cannot remain 
for ever without change.' Cf. the ???????? ???????? ????? at 283d. 
272 
Necessity 
the eyes—parts of a body made from the four elements—and the 
behaviour of light-rays. This physical mechanism of sight Plato  
describes in detail D5b-46c), calling it and other organs and processes 
'co-causes',1 secondary and subordinate. The error of most earlier 
philosophers had been to regard them as primary. In the Phaedo he had 
castigated this neglect of final causation as 'absurd' and 'sheer laziness', 
and dismissed as waste of time the attempt to explain the world by 'airs, 
ethers, waters and other strange things'.2 Now his attitude has changed. 
Under the title 'what happens of necessity', material conditions and 
processes occupy at least a third of the whole work, and detailed 
explanations are given of the ingenuity with which the Demiurge 
adapted them to good ends. The cosmos is 'the combined work of 
Reason and Necessity', Reason prevailing over Necessity, in Reason's 
way, by 'wise persuasion' D76-48a). The personification of Necessity 
as 'persuadable' is apt. In Greek poetry, including the philosophical 
poets Parmenides and Empedocles, the goddess Ananke, true to her 
name, was inexorable, and ' of unconquerable might'.3 Plato corrects 
this in its own mythical terms: for the most part she has yielded to the 
arguments of Reason. 
But not entirely. In turning to the negative aspect of necessity, Plato 
drops the personification, which has served its allusive purpose.  
Necessity, 'the errant cause' D837), is of the kind which 'destitute of reason 
produces chance results, without order'. They are not designed, but 
just happen D6 e, 47 e). Matter has its necessary characteristics (' powers') 
indifferent to reason or values. Fire may warm a house and cook 
a meal, or destroy the house and kill its owners. The latter we call 
an accident, and it explains the close association of necessity and chance 
in the Greek mind.4 The necessity is internal to one thing, as heat to 
fire, the chance lies in the proximity of two things, burning agent and 
combustible material. Although fire must burn, the craftsman by his 
1 ???????? 46 c, the word used in Pol. of ?????? subordinate to a major one, as the manufacture 
of spindles and shuttles subserves the art of weaving B81 c-e). 
2 Pho. 98b~99c; vol. iv, 330, 350. Contrast esp. 77m. 68e-69a, translated on p. 252 above. 
3 Aesch. PV 105, Eur. Ale. 965. For Ananke in Greek literature and Parmenides see vol. 11, 
34-7; in Empedocles, ib. 163. 
4 For necessity and chance as practically identical for the Greeks, see vol. 11, 414 f., and cf. 
Cornford, PC 165 ff. In Plato it is best illustrated by the atheists' account of the fortuitous origin 
of the world at Laws 889 a-c, esp. in the phrase ???? ???? ? ?? ??????? (ci). 
19 
273 
GHG 
Timaeus and Critias 
choice of materials and design can reduce the danger of accidents and 
direct its activity, so far as possible, to useful ends, thus 'persuading 
Necessity \ She symbolizes the ultimate intractability of matter, which 
no craftsman can overcome entirely. The cosmos is a magnificent 
creation, modelled on the Forms by divine Reason, but being corporeal 
it cannot be the Forms, any more than a block of marble can be— 
though it may be made to resemble—a human face. Pygmalion is a 
myth, and even God could only make the world 'as good as possible', 
'to the best of his powers'.1 
The central idea of a material chaos moved by a mindless inner 
necessity bears a resemblance to the system of Democritus which can 
hardly be fortuitous.2 'Democritus', said Aristotle, 'ignored the final 
cause and attributed all the operations of nature to necessity.' According 
to Diogenes Laertius he held that 'All things come about of necessity, 
for the vortex is the cause of all becoming, and this he calls necessity', 
and the Placita have it that necessity for him consisted in ' the resistance, 
movement and blows of matter'.3 For Democritus this was all that was 
required to produce our world. In Plato's eyes its beauty, goodness and 
order could never have emerged from such a welter without an  
intelligent disposer. He therefore speaks of it as the state of things 'before the 
creation of the cosmos'. Hence his duality of primary and secondary 
causes, which whatever its difficulties does avoid the intellectual dilemma 
facing those who try to reconcile an omnipotent and benevolent deity 
with the manifest imperfections of the world. 
1 This is repeatedly emphasized: 3033, 37d2, 46c8, 53b5. 
* Few statements about Tim. are uncontroversial. Said Taylor in 1926 (Comm. 3): ? believe 
I shall be able to show... that there are no traces anywhere in Plato of a knowledge of Democritus, 
and that in the Timaeus in particular the whole plan of the dialogue makes such references 
impossible.' Contrast Jowett's editors {Dialogues vol. m, 1953, 669 n.): 'Most authorities would 
now agree that the Timaeus is in part directed against Democritus.' This is true, so there 
is no need to quote individuals. Many refer to the study of Hammer-Jensen in AGP 1910, who 
weakened her case unnecessarily by the unacceptable claim that P. only learned of Democritus's 
work when 77m. was partly written, and changed his mind in the middle. (For some criticism 
see vol. 11, 406 n. 2.) Stenzel in his essays on P. and Democritus tends to exaggerate 
resemblances between them, but is helpful nevertheless. 
3 Arist. GA 789b!, D.L. 9.45, Aet. 1.26.2 (DK 68 a 66 and 1). 
274 
Creation of cosmos 
The creation of cosmos 
Why it was created B9d~3ob). This section is the starting-point of 
Timaeus's narrative. The reason for creation is religious. God is good, 
and it is not * lawful' (or permissible, themis) for the Best to act  
otherwise than for the best. Being good, he had no jealousy in his nature (a 
criticism of current ideas of divine phthonos). He wished everything to 
be as good as possible, and so, finding visible nature in its state of  
restless and inharmonious motion, reduced it to order, which he deemed 
better than disorder. Reasoning further that anything in nature would 
be better with intelligence than without, and cannot have intelligence 
without soul, he put mind in soul and soul in body.1 So by God's 
providence this world was created as a living and intelligent creature, 
modelled on the intelligible and all-embracing Living Creature itself. 
Uniqueness of the cosmos Cia-b). Democritus had argued that since 
there exists an infinite number of atoms moving at random in an infinite 
void, it was unreasonable to suppose that the chance collisions which 
have produced a cosmic system in our part of the void would not have 
led to the formation of similar systems elsewhere.2 What is at stake 
therefore is not only the number of worlds, but the fundamental  
question of chance or design as the originating principle, and Plato's reply 
is in terms of purpose: 'In order that it might resemble ...' The cosmos 
must resemble its model in every way possible, the model is unique, 
therefore the cosmos must be unique. How the uniqueness of a physical 
world can be ensured is explained later C2c~33a, p. 279 below). 
The case for the uniqueness of the model is more complex. Plato has 
argued in different ways in different dialogues that every form is unique, 
but the present argument is tailored to the Form in question, namely 
the Form of Animal, and is perhaps not exactly like any of the  
arguments for the uniqueness of a Form in general. There are in fact three 
1 Where nous and psyche are distinguished, psyche signifies what for P. are the lower parts of 
the tripartite soul, the life of an animate body with the capacity for sensation, desire and emotions 
such as anger and fear. Although it would be nonsense to say that God is not alive, I think, as I 
have said on p. 215, that in the context oj the distinction God, having no body, has nous without 
psyche. It is ?? ???? ????? ????? that cannot have one without the other C0 bi). 
2 Some at least, he thought, would not repeat the features of our world exactly. For innumerable 
worlds in Democritus see vol. 11, 405, 406 n. 1. 
275 
19-2 
Timaeus and Critias 
arguments which have sometimes been taken as one and the same,1 and 
have generated much comment. They are, besides the present one, the 
argument about the 'three beds' at Rep. 597c and the 'Third Man 
argument' of the Parmenides. I have tried to disentangle these in vol. iv, 
552, and will only repeat here the gist of the one which now concerns us, 
at Tim. 31a. It runs as follows. 
The Form of a genus (in this case Animal) must contain the Forms of 
all the species subsumed under it. In Plato's terms, they are 'parts' of 
it.2 If there were two, each would contain some only of the relevant 
species, and there would have to be a more all-embracing Form  
containing both of these with the species which each embraces. They would 
be like the Forms of vertebrates and invertebrates, each containing a 
large number of species of animal but not all. 
The body of the cosmos C1 b-34b). The early Ionians had assumed one 
primary substance in the cosmos, Parmenides two,3 and Empedocles 
four, fire, water, earth and air. Plato agrees with Empedocles, but unlike 
him offers reasons. Like everything generated, the cosmos must have 
body and be visible and tangible (solid). This demands two bodies, fire 
(which includes light, 45 b) and earth. So much the Demiurge must 
accept, but now he gets to work himself. Two things cannot be 'well' 
(kalds) combined without a third to bind them together. This is because 
the strongest bond is geometrical proportion, which cannot exist  
between fewer than three constituents. The fourth is added because the 
cosmos is to be three-dimensional, and whereas for a plane surface (i.e. 
to link two numbers) one mean proportional suffices, a solid body 
requires two.4 We begin to see what Timaeus meant when he said that 
1 E.g. by Adam in his note on Rep. 597 c. 
2 ???? 3i a6. Cf. p. 153 n. 4 above. 
3 In the 'Way of Seeming', of course. In fr. 8.53 the two are light and night, but he seems in 
some way to have identified night with earth. Aristotle says four times that his two primary 
substances were fire and earth, Theophr. repeats it, and Alex. Aphrod. adds explicitly that he 
called the earth darkness. (See vol. 11, 58 with n. 2.) These commentators had the complete poem, 
and even in the surviving frr. {I.e.) he calls night 'dense and heavy'. Plato's start from fire and 
earth may reflect once again his great respect for Parmenides. 
4 P. is speaking in terms of square (or rectangular) and solid numbers (Taylor, Comm. 97 f.). 
His last statement is not universally true. See Grote, PL m, 252 n. a, Taylor, Comm. 97f. and 
on these mathematical passages in general Cornford, PC 45-52 (with his quotations from Heath) 
and Archer-Hind, Tim. 97-9. 
276 
Creation of cosmos 
his listeners would follow his demonstrations because they were 
experts, trained in the requisite sciences E3c). 'Plato is compressing 
his statement of technical matters to such an extent that only expert 
readers would fully appreciate his meaning' (Cornford, PC 47). Apart 
from that, his Pythagorizing synthesis of the mathematical and the 
physical may seem strange today.1 If a medicine which requires  
ingredients mixed in the proportion 1:3:7 is made up in the proportion 
1:10:100, it will not console us to be told that the latter is a perfect 
geometrical ratio. It was too much for Aristotle's common sense. ' One 
might also ask', he protests (Metaph. 1092b26), 'what good things get 
from numbers by their mixture being in accordance with a number . . . 
Honey-water is no more wholesome if it is mixed in the proportion of 
three times three: it would do more good if it were in no particular ratio 
but well diluted than if it were numerically expressible but strong.' Nor 
is it obvious why two elements (water and earth, say, to make clay) 
should need a separate third to 'bind' them. Cornford says nothing of 
this, and Taylor (Comm. 95) treats it as a special condition for earth and 
fire: they need a mediating element because their own characters are so 
strongly opposed. This will not do. The condition, laid down at 
38b8-ci, is purely general: 'It is impossible for two things to be well 
combined without a third.' Plato is using the language of mathematics, 
not of chemistry or any science which takes account of the physical 
properties of different kinds of body. 
The Politicus and Philebus2 have accustomed us already to the  
importance in Plato's eyes of measure, limit and proportion as essentials of 
goodness (fitness for function). He does not say that two elements by 
themselves cannot mix: the emphasis is on kalos. The cosmos, though 
not perfect, is the best and most lasting of all created living things. It 
cannot therefore have been thrown together haphazard, but was 
planned as an organism in which the various components are blended 
1 For the Pythagorean derivation of bodies from geometrical solids, and ultimately from 
numbers, see vol. i, ch. iv(d), esp. pp. 229-73. 
2 Not to mention Gorg. 5076-508 a, which applied the laws of mathematical proportion to 
cosmic structure and human conduct alike. The same association of them with values permeates 
the Rep., but it reaches its climax in 77m., both here and later in the construction of the geometrical 
particles. 'Of course', says Taylor of the present passage (Comm. 98), 'this is not given by Plato 
as a demonstration' that there are exactly four ' roots'.' It is simply a play of mathematical fancy.' 
I suspect that for P. it was more than that. 
277 
Timaeus and Critias 
with the most exquisite delicacy and precision. This proportionate 
blending ensures its wholeness and unity C2d9-33a 1), knitting its 
parts together in bonds of amity1 indissoluble save by its author. In the 
Phaedo Socrates demanded an explanation of the world which would 
demonstrate that what binds and holds it together is the power of the 
good and right.2 Here where Plato gives the full answer, the binding 
force is expressed in terms of analogia, geometrical proportion. No 
reader of the Gorgias and Republic will be surprised. 
God, then, made the elements 'as far as possible proportionate to 
each other, so that as fire is to air, so air is to water, and as air to water, 
so water is to earth' C2b). This is usually, as by Cornford,3 referred to 
their respective quantities, but with the description of pre-cosmic chaos 
fresh in our mind, some questions must suggest themselves, perhaps  
unfairly, to the literal-minded. The Receptacle of Becoming itself had a 
fiery part, a watery part and so on, and it was also said that fire, water, 
earth and air, though 'without proportion and measure', already  
possessed traces of their distinct natures. Those four, and no others, were 
data. The Demiurge only imposed order on them through number 
E3a-b). We are about to learn also that, for excellent reasons, he  
incorporated all the fire, water, air and earth into the cosmos. How then, 
first, was it in his choice to have, for mathematical leasons, four elements 
rather than two, and how, secondly, if he used all there was of them, 
making no selection, was it open to him to relate the quantities in 
geometrical proportion?4 One might also wonder, thirdly, since the 
creation is cast in narrative form, whether we are to imagine that he has 
already organized each element, whose quantities are here decided, into 
1 P. uses ????? C202), the Empedoclean term for the unifying force. Empedocles, a westerner 
like Timaeus, also introduced the idea of definite numerical (if not geometrical) proportions, at 
least for organic compounds, though in a less advanced way and without P.'s teleological 
implications. See vol. 11, 211-16. 
2 99 c. English cannot reproduce the affinity of the Greek words in ?? ???? ???????. 
3 PC 43: 'All that the Demiurge does now is to fix their quantities in a certain definite 
proportion.' Cf. 51:' Plato has not indicated what are the quantities between which his  
geometrical proportion holds ... It may be conjectured that the quantities in question are the total 
volumes of the four primary bodies.' 
4 I might mention R. J. Mortley's alternative suggestion, in a note in Hermes 1969, that 
'The numbers that could be used in a proportion such as this would not represent relations 
between elements, but cosmic forces existing as Forms and affecting the sensible world in 
the same way as other Forms.' I do not feel able to comment on this, but some may find it 
helpful. 
278 
Creation of cosmos 
minute corpuscles of geometrical shape.1 This is not mentioned until 
much later, after the description of pre-cosmic chaos, where it certainly 
appears as his first step in the introduction of order. It seems possible 
that the randomness of Timaeus's discourse and his repeated fresh 
starts, which he puts down to human weakness, besides conveying an 
air of spontaneity, have the ulterior purpose of making these  
discrepancies less noticeable. Certainly they do not seem to have occurred 
to previous commentators, who have followed Plato's order of  
exposition rather than what he himself says is the right one. If so, one can only 
sympathize and agree with his reasonable plea that on a subject like the 
origin of the whole universe one should not expect an account in every 
way self-consistent and precise B8 c). 
To make the body of the cosmos, the Demiurge used the whole 
quantity of all four primary bodies C2c), thus ensuring first its unity 
(there was nothing left over from which a second might arise, 33 a 1) 
and secondly its permanence. This has already been attributed to the 
4amity' brought about by the proportions between its elements, but 
Timaeus now adds that the only causes of sickness and senility in living 
creatures are attacks from heat, cold and other 'powers' of extraneous 
bodies.2 Immune from these dangers, although material it can last as 
long as God wills; and since God himself says later D1b) that only an 
evil being would wish to dissolve what is good and well constructed, 
it will last for ever; but its preservation lies in God's hands, not in its 
own nature, for nothing bodily can be intrinsically indestructible. 
Certain other consequences flow from its completeness, and differentiate 
it from any of the living creatures which it contains. It needs none 
of their organs or limbs, for there is nothing outside for it to see 
or hear, nothing to eat or excrete—it is entirely self-sufficient—or 
grasp with hands, nowhere to go on legs and feet. So he made it a 
sphere, the best of all shapes, which contains all others3 as the cosmos 
1 If, as Cornford seems to have thought {PC 223), the mention of numbers (?????) in the 
summing-up of God's work on the corpuscles at 56 c, refers to the relative quantities of each kind, 
this would seem to be so. 
3 In case we should fail to take the animation of the cosmos seriously, this is an effective 
reminder of its affinity to the rest of animal life. 'Analogy' would be too weak a word. 
3 It is the only figure in which all five regular polyhedra can be inscribed, and these underlie 
the structure of the primary bodies (pp. 282 ff. below). Inscription is mentioned at 55 a. This is 
given by Proclus as the more probable of two explanations {In Tim. 2, 71 and 76 Diehl). 
279 
Timaeus and Critias 
contains all living creatures, and gave it as sole motion revolution 
about its axis, the only motion which a body can perform within 
its own limits. 
In this, and especially its psychological connotations, Plato owes 
something to Alcmaeon, but more to Parmenides, especially in the 
denial of some features of Pythagorean thought with which they were 
both familiar. He takes over the One Being of Parmenides,' complete', 
Mike the mass of a well-rounded sphere', 'equal every way from the 
centre' (fr. 8.42-4), with nothing outside it.1 Even when he allows for 
Becoming by restoring motion and heterogeneity, he continues to 
respect Parmenides's next dictum, that it 'keeps evenly within its limits' 
(v. 49). Nor can Empedocles have been far from his thoughts, who  
describing the sum of things in the reign of Love wrote (fr. 29):' No twin 
branches spring from its back, there are no feet nor nimble knees, but 
it was a sphere and in all directions equal to itself.' He even  
contributed the conception of a great Mind pervading the whole cosmos (fr. 
134). Cosmogonical theory is not Plato's metier, and he does not  
hesitate to hark back to earlier leaders in the field if he can adapt them to his 
demonstration that the world is born of design not chance; for on that, 
as he makes even clearer in Laws 10, depended the existence of  
objective criteria for human behaviour. He regrets therefore the view  
common to the Milesians and Democritus that the cosmos is surrounded by 
a mass of bodily substance out of which it had arisen and into which it 
would some time disintegrate; for that was linked to the conception of 
it as a product of mindless physical forces alone. 
Construction of the primary bodies E3c—57c!) 
So far Timaeus has described the body of the cosmos in fairly general 
terms, apologizing at the same time for not treating first of the soul, 
which was created first. The heavenly bodies, on the other hand, are 
1 For Alcmaeon see vol. i, 354; for Parm. vol. 11, 47 f. Plato even recalls his language. Cf. 
Tim. 33 b 4-6 ???????????, ?? ????? ?????? ???? ??? ???????? ???? ?????? ?????????? ?? ???? ????? 
???????? with Parm. fr. 8.42-4 (^vhich he quotes verbatim at Soph. 244?): 
???????????? ???? 
???????? ???????? ??????? ?????????? ????, 
???????? ???????? ?????. 
Mortley has drawn attention to the meaning of ?????????? and its connexion with Parm. in an 
article on Plato's choice of the sphere in REG 1969. 
280 
Construction of the primary bodies 
displaced from the account of the world's body and brought in after its 
soul in explanation of time, which depends on their revolutions C7dff., 
pp. 299 below). His order has indeed something of a casual air,1 and 
just as he described the creation of cosmos before the pre-existing chaos, 
so he defers the structure of the elements of which the body of the 
cosmos consists until after the account of its soul, of time, the heavenly 
bodies, and even the creation and destiny of man. It arises, naturally 
enough, out of the description of the pre-cosmic chaos, in which 
4 traces' of the elements tossed about' without proportion or measure' 
(p. 269 above). To reduce them to order, the Demiurge ' fashioned 
them by shapes and numbers'. We have already examined their  
previous state, and since they are the constituents of the world's body 
Cib~32c) their formation may best be taken here. 
The geometrical basis of matter. If the world is the work of reason, 
rationality (displayed in measure and proportion) must be detectable in 
the ultimate and most elemental forms of which it is built up; and these, 
he has warned us D8b-c), are not simply the earth, water, air and fire of 
Empedocles. Those are more complex than syllables, and to find the 
actual ' letters' of the universe one must probe deeper. The atoms of 
Democritus were of all kinds of irregular shapes and sizes, which 
suited his general view of the world as a product of undesigned  
coincidence. With an infinite number of irregularly shaped atoms colliding 
and becoming entangled in infinite space, it was inevitable (ananke) that 
somewhere, some time, they should build up into a world like ours. 
Implacably opposed to such a view of its origin, Plato was bound to 
carry his opposition into the ultimate structure of matter and show that 
even an atomic theory need not be atheistic. Against the Democritean 
jumble he set the Pythagorean idea that number and measure entered 
into everything. The Pythagoreans, said Aristotlfe (Metaph. 985b 32, 
and similarly in many places), since the nature of everything else 
seemed to be entirely assimilated to numbers, and numbers to be  
primary throughout the natural world, supposed the elements of numbers 
1 I hope this will not be taken to imply anything but the most artful composition on Plato's 
part, appearing not least in the impression he conveys of an expert giving an impromptu talk to 
friends rather than a formal lecture. 
28l 
Timaeus and Critias 
to be the elements of all that exists, and the whole universe to be a 
harmonia and a number.1 
The details, as Timaeus says, are for mathematicians. The general 
scheme is based on the five regular solids or polyhedra: tetrahedron 
(three-sided pyramid), cube, octahedron, icosahedron, dodecahedron.2 
Their regularity, and doubtless also the fact that they can all be  
inscribed in a sphere, 3 gave them, at least in Pythagorizing eyes, a 
peculiar perfection and beauty. Plato calls the first four 'the four 
surpassingly beautiful kinds of body' E367). The pyramid, as the 
smallest, most mobile and sharpest of these regular solids,4 is assigned 
to fire, the icosahedron to air, octahedron to water and cube to earth. 
These shapes and sizes are connected with their physical qualities, the 
destructive power of fire, the stability of earth and so on. So far the 
theory may be called particulate if not atomic, for the solids are bodies 
too small to be seen, though visible in the aggregate E6b-c). As bodies 
they cannot be geometrically perfect, but have been made by God as 
accurate as the nature of Necessity would allow. They are not strictly 
atomic (indivisible), for they can be actually divided (not merely 
analysed by the philosopher) into yet more elementary forms. One of 
the objects of Plato's theory of matter was to allow for the mutual 
transformation of certain elements. Both he and Aristotle rejected the 
theory of Empedocles who, under the direct influence of Parmenides, 
had denied the change of any of the four 'roots' into another: all 
phenomena were accounted for by their mingling and separation (fr. 
21.13-14; vol. 11, 148). In combination they retained their individual 
identities, though these might be imperceptible. 
1 Some scholars who deny that P. had Democritus in mind point as evidence to the fact that 
their theories are quite unlike in this respect, that P.'s atoms are geometrically constructed— 
as if any reference of P. to Democritus could be other than polemical. 
3 Owing to the fame of Tim., the five regular polyhedra on which P. based his atomism became 
known as 'the Platonic figures', but he was certainly not responsible for their recognition, and the 
evidence that their connexion with the four elements and the cosmos as a whole was already a  
feature of Pythagoreanism is strong. It is fully discussed in vol. i, 266-73 (together with the date of the 
construction of the five regular polyhedra), and cf. the quotation from von Weizsacker on pp. 225 f. 
3 Perhaps also the beauty of crystals. Ridgeway observed in CR 1896 that quartz crystals are 
pyramidal, iron pyrites cubical, and garnets dodecahedral. (There is a reference to inscription at 
55 a3, and it is probably also implied at 33b3-4.) 
4 A glance at scale models (of which I have a set in front of me as I write) or drawings brings 
home the relative smallness and sharpness of the pyramid, assuming (as one must) that all the 
figures have sides of the same length. (The drawings on p. 76 of Lee's translation are helpful but 
not to scale.) 
282 
Construction of the primary bodies 
Transformation of the primary bodies. To combat this view, Plato carries 
his analysis a stage further. The surfaces of three of the four polyhedra 
which he has assigned to the elements are triangular, and the square 
itself can be divided into two triangles.1 For reasons not immediately 
obvious,2 Plato does not take the equilateral triangular faces of the first 
three as ultimate, but divides them, as well as the square, into two 
right-angled triangles, scalene and isosceles respectively. These he posits 
as truly elementary, all other triangles being derived from them E3 c-d). 
Since, then, the particles of three of the elements have identical surfaces, 
it is possible that, if they should be broken up, the surfaces will re- 
combine in different ways to form any other of the three regular solids 
so constructed. When for instance heat dries up a puddle of water, the 
small, sharp, mobile pyramids of fire have pierced and split up the water- 
particles, and the twenty faces of each have regrouped themselves as 
two octahedra (air-particles) and one pyramid (fire) E6c!). Earth 
alone is not subject to this process of transformation, for its surfaces 
can only be reduced to isosceles triangles which cannot combine with 
the others. If earth is split by fire, its parts simply drift about until they 
meet others of their kind and can re-combine as earth.3 
1 Cornford (PC 211) said it is 'by no means obvious* why Plato does not take the square as 
one of his elementary plane figures. Others have also seen it as a problem. Surely the reason is 
that it is not an elementary plane figure. It can be analysed into triangles, but with triangles the 
analysis of a rectilinear plane figure must stop, as Plato noted E3 C7). 
2 Cornford B12) attributed it simply to the choice of the regular solids. We cannot here 
pursue Plato's geometrical scheme through all its details, but mention must be made of Popper's 
brilliantly argued thesis that the special importance of these triangles lies in their incorporation 
of the irrational square roots of 2 and 3. Plato's main contribution to science, in Popper's view, 
sprang from his realization of the problem of the irrational (on which he lays such stress at Laws 
820 a-b), and his consequent substitution of a geometrical view of the world for the arithmetical 
outlook of early Pythagoreanism. See Popper's C. andR. 75-93 (rePr· m Brown's Meno 143-73) 
and O.S. 1, ch. 6 n. 9, 248-53. The thesis is summarized by Toulmin and Goodfield, A. of M. 80 
(in both original and Penguin eds.: pp. 75-82 of this book give a lucid summary and appraisal of 
Plato's theory of the composition of matter). I was puzzled at first that Plato should introduce 
irrationals at the very heart of his scheme, when he has been so emphatic (here and in Pol. and 
Phil.) that the work of Mind was always characterized by measure and ratio (????? ? and ????*). 
But Popper's explanation of how, precisely because of this innovation, 'the existence of irrationals 
was no longer incomprehensible or "irrational"' has removed this difficulty. See O.S. 1 at top of 
P-25I. 
3 56d. Aristotle, who believed that transmutation occurred between all four simple bodies, 
complained that in making earth an exception P. was exalting his own mathematical theory at the 
expense of observed facts (Cael. 30635-9). This criticism has been repeated in modern times. 
Cornford (PC 216) thought it 'simply a consequence of the decision to assign the cube to earth', 
and Eva Sachs surmised that P. would have been happy if someone could have supplied a fourth 
283 
Timaeus and Critias 
Fifth figure and fifth body. Of the five regular solids there remains the 
dodecahedron, whose surfaces are pentagonal. It was used by the 
Demiurge not for any of the simple bodies, but for the whole cosmos. 
This, as we know, was spherical, and commentators since Plutarch have 
aptly compared Phaedo nob, where the spherical earth, seen from 
above, is said to resemble ' a ball (??????) made of twelve pieces of 
leather' and to be 'picked out in various colours'.1 This raises an 
interesting question in the history of ideas, namely the emergence of 
the notion of a fifth element, Aristotle's aither. Each of the other poly- 
hedra is associated with a simple body. The dodecahedron is not. The 
heavenly bodies are said to be largely made of fire, while aither is 
simply the clearest form of air D0a and 58d). But the symmetry of a 
scheme in which all five figures corresponded to different simple bodies 
must have made a strong appeal,2 and it is about this time, or a little 
earlier among the Pythagoreans, that the first hints of a fifth body 
appear. It is unambiguously present in the Epinomis (981c), and even 
in the Cratylus aither is expressly separated from air Diob).3 Other 
evidence has been examined in vol. i,4 where it is suggested that the 
regular body made up of triangles similar to those of the others. Others agree (Solmsen, ASP1V 
52 n. 124; G. E. R. Lloyd, EGSc 77), but Proclus (ap. Simpl. Cael. 643) defended P.: earth is 
never seen to change, though earthy compounds do when water or fire leaves them. In modern 
times Cherniss (ACPA 150) and Taylor incline to the empirical explanation: 'It is because of the 
irreducibility of earth that Timaeus needs two primary triangles' (Comm. 369). Cornford himself 
said earlier B13) that P. wished to explain transmutation, and 'for this physicalpurpose all he needs 
is triangles which can be reformed into solids of a different pattern* (my italics). Though there 
can be no proof either way, this seems to me the more likely order of P.'s thought. If so, the 
words cbs ???????? and obs ???????? at 49b8 and C7 are to be taken seriously, which is reasonable. 
Cf. ????????, ?????????? at 54 b. 
1 ???????, ??????? ???????????. Burnet saw in these words the explanation of ???3??????? 
at 5 5 c6, denying that it could refer to the signs of the Zodiac as usually supposed {EG? 294 n. 5). 
Cf. Cornford, PC 219: 'not only the twelve signs of the Zodiac, but all the other constellations'. 
On the sphere constructed out of twelve pentagons see further vol. 1, 268 f. 
3 This is how Simplicius saw it (Phys. 1165,18): 'Why then does [Aristotle] calls the heavens 
a fifth body? Perhaps because Plato himself describes the substance of the heavens as different 
from the four sublunary bodies. After all, he assigned the dodecahedron to the heavens, and each 
of the four elements he described by a different figure.' 
3 In Pho. too, ? ?? ???? ???, ???????* ??? ?????? occurs in the cosmological myth (nibi). 
The fifth body was ready and waiting, so to speak, in common belief and in myth, to be adopted 
into natural philosophy. 
4 Pp. 267-73. To the modern references given there may be added P. Moraux, art. ' Quinta 
essentia* in RE xlvii. Halbb. 1171-1263 with Nachtrag 1430-2; Harward, Epin. 125f.; Taran, 
AJP 1962, 315f. (reviewing Novotn^'s Epin.), where the statement that Xenocrates 'located 
[aither] outside the sphere of fire* must be an inference from the order in which X. mentioned the 
five bodies. (In his Academica of 1975, p. 40, Taran concludes from Simpl.*s testimony that 
284 
Construction of the primary bodies 
conception of a fifth element evolved gradually out of earlier cosmo- 
logical presuppositions. 
The remoter principles: geometry and physics. At 53 d Plato hints that 
even his analysis of physical bodies into triangular surfaces is not  
complete: 'The principles more remote than these are known to God and 
such men as he favours.' Being concerned with the physical world, the 
Timaeus has no need to go further back than the surface, which as the 
boundary of the third dimension, depth, makes sensible body possible.1 
But the Pythagorean flavour of the whole makes it easy to guess what 
these remoter principles are. First come Limit and the Unlimited, equated 
with numerical oddness and evenness. They produce the unit, first  
imposition of Limit on the Unlimited, from which spring numbers. From 
numbers are derived geometrical figures by equating the unit with the 
point, two with the line, three with the simplest rectilinear plane figure. 
From plane figures come solids, and from solids sensible bodies.2  
Aristotle never tired of castigating the Pythagoreans for this derivation of the 
physical—visible, tangible—from mathematical abstractions. 'They 
assumed the principles of mathematics to be the principles of  
everything. '' They suppose units to possess magnitude.'' When they construct 
physical bodies out of number—things that have lightness and weight out 
of elements which have neither—they appear to be talking about some 
other universe and other bodies, not those that we perceive. '3 In De 
caelo B99 a 1-300 a 19) he levels similar criticisms at the Timaeus itself. 
As so often, opinions differ. Thus Cornford, PC 285: 'We must 
reject the view that Plato has reduced the bodily to mere empty space 
figured in the geometrical patterns which the Demiurge is now going to 
introduce'; but Burnet, T. to P. 344: 'Plato undoubtedly means to say 
that the corporeal can be completely reduced to extension geometrically 
limited.' The question is obviously bound up with the nature of the 
'Receptacle of becoming', that 'dim and difficult' something which 
Xenocrates 'identified the ether with the dodecahedron', i.e. believed that this was Plato's  
intention.) 
1 53 c, A. T. Nicol in CQ 1936, 125. 
3 Alex. Polyhist. ap. D.L. 8.24. For a full account of the Pythagorean theory, with authorities, 
see vol. 1, 238ff. For Plato, Stenzel's discussion in Z. und G. 70-5, 'Das Ende der Teilung des 
Raumlichen im math.-physikalischem Atom', is relevant. 
3 For reff. and further quotations see vol. i, 229, 232, 234, 235. 
285 
Timaeus and Critias 
neither senses nor mind can properly comprehend. The difficulty for 
Plato results, I suggest, from an attempt to reconcile two different types 
of cosmology: the Pythagorean, predominantly mathematical and  
paying the minimum of attention to physical substances or properties like 
fire or earth, hot, cold, wet and dry; and the Ionian or materialistic, 
culminating in the Heraclitean conception1 of the world as a never- 
ending flux of change. Hot is continuously becoming cold and cold hot, 
water is drying up and turning to air, air condensing and becoming 
water, with never an instant's pause. We know what a tremendous  
impression this world-view made on Plato, but if absolute instability 
represented the true nature of the universe it could never be the object 
of scientific knowledge, for it could never be brought under general 
laws. When he wrote the Phaedo and Republic he appears to have 
accepted its consequences and abandoned hope of a science of the 
physical world. Knowledge is not of'what becomes' but only of'what 
is', the immutable world of Forms after the pattern of which the  
temporal world is formed. The Theaetetus probed the question further on the 
epistemological side, and now comes the Timaeus. Here he maintains as 
strongly as ever the distinction between 'what becomes' and 'what is', 
with its parallel epistemological distinction between belief and  
knowledge, and warns that any account of the physical world can be no more 
than probable; yet such an account has now become for him supremely 
worth giving, and he takes great pains to work it out in detail. His  
conclusion seems to be this. If Heraclitean flux or Democritean atomism has 
the last word, the world in which we live must be abandoned to chance. 
No other cause brought it into being or sustains it now. But this belief 
is both erroneous and (as he will argue at length in the Laws) morally 
disastrous. True, the creative Reason had to work on a given and to 
some extent recalcitrant material. Perfection is found only among the 
Forms, not in space at all; but even in this world Reason has overcome 
Necessity to a large extent, and the study of mathematics and above all 
of astronomy will quickly satisfy a thoughtful man that the primary 
impulse behind the creation of the universe is rational.2 
1 The conception of contemporary Heracliteans rather than of Heraclitus himself (p. 80 
above). 
2 Cf. Phil. 28 e, p. 214 above. The part of astronomy in the argument from design, which 
becomes one of the main themes of 77m., was echoed in remarkably similar terms by Newton in 
286 
Construction of the primary bodies 
If, then, the cosmogony of the Timaeus, represented by Plato as the 
conquest of (Democritean) Necessity by Reason, may be crudely 
described as an attempt to impose the Pythagorean mathematical scheme 
of reality on the Heraclitean flux of becoming, we need not be surprised 
to find it less than wholly successful; and this is suggested by his 
hesitant and fumbling description of the Receptacle. It is better to 
appreciate his state of mind, and the philosophical situation which gave 
rise to it, than to try to force into clarity and consistency what was for 
its author himself 'dim, difficult, hardly to be believed'. We have seen 
that the rest of his description forbids us to think of it as mere empty 
space. It might be so if one could abstract from it the 'motions' and 
'powers' which continually surge about in it, but not only does it 
never exist without them: it is inconceivable without them, for they are 
in a sense qualities of itself. True, it must be imagined as per se qualiti- 
less, like the oil which must be odourless to serve as a base for perfumes. 
Yet, says Plato E1b), one should not speak of fire, water and the rest 
being in it, so much as of the incandescent part of it, the liquefied part of 
it and so on. Even before the ordering by figures and numbers began, 
it contained inchoate forms or 'traces' of the four simple bodies. 
Physical matter, solid substance, did exist, but 'without proportion or 
measure', in fact as an unlimited awaiting the stamp of Limit. 
The most probable conclusion is that the particles created by God's 
conversion of chaos into cosmos were genuine corpuscula in the shape, 
so far as Necessity allowed E6c 5), of one or another of the regular 
polyhedra, which gave each of the popular elements its character as 
fiery, wet and so on. Unlike the traditional Pythagoreans, Plato could 
distinguish when he liked between mathematical figures and their 
approximations in material objects, models or drawings {Rep. 510c, 
Phil. 62a-b). Yet in this his most Pythagorean dialogue1 he does not 
always observe the distinction. In more than one place he seems to 
the Principia (ref. in Cornford, Princ. Sap. 21): 'It is not to be conceived that mere mechanical 
causes could give birth to so many regular motions . . . This most beautiful system of the sun, 
planets and comets, could only proceed from the counsel and dominion of an intelligent and 
powerful Being.' 'Counsel and dominion* almost translates Tim. 48a2 voO . . . ???????? ?? 
?????? ?. 
1 Though like everyone else I cannot accept Taylor's thesis that in Tim. P. does not give his 
own doctrine but only a historical account of fifth-cent. Pythagoreanism, the fact that so 
experienced a scholar could hold such a belief says much about the character of the dialogue. 
287 
Timaeus and Critias 
assume that once a geometrical construction has reached the third 
dimension, one has immediately a perceptible body. 'Every kind of 
body has depth, depth includes surface, and every rectilinear surface is 
composed of triangles' E3c). Conversely at 53e~54a 'the four most 
beautiful bodies' appears to refer to the figures, including the triangles 
of which they are constructed. In the Laws he is even more definite 
(894 a): 'What is the condition for the coming-to-be of all things? It 
occurs when a starting-point is extended to the second dimension and 
thence to the next, and having achieved three dimensions becomes 
perceptible to whatever has senses.'1 This is pure Pythagorean doctrine 
as described by Alexander Polyhistor (p. 285 above) and criticized by 
Aristotle. It bears on a problem that has never been solved.2 How can 
triangles float about by themselves, as the triangles of a disintegrated 
earth-particle are said to do, until they can re-unite with their own kind? 
Martin's conception of them as 'thin plates of corporeal matter'  
enclosing empty space {Tim. 11, 241 f.) has found little favour.3 The reverse is 
surely correct, that the triangles are surfaces bounding solid corpuscles, 
and so giving geometrical form to previously formless matter. This 
does not solve the problem of the drifting triangles. I would dare to say 
that it did not present itself to the Pythagorizing Plato of the Timaeus. 
Particles vary in si{e. At 54c!—5 5 c the faces of three of the polyhedra are 
said to be further divided into six triangles instead of two, and those of 
the cube into four. No plausible explanation of this was forthcoming 
until Cornford {PC 234^) connected it with the statement at 57c-d 
that the triangles exist in various grades of size, and that this explains 
the varieties to be found in each element. As Timaeus goes on to say, 
1 For the language (???? = dimension) cf. Rep. 528b, which gives the cube as an example of 
the three-dimensional, and Epin. 990d ???? Tpis ?????????? [????????] ??? -rfj ?????? ????? 
???????. At 99°a ?????? ? is equated with ?????. 
2 Unless it is a solution to say with Cornford (PC 229f.) that it 'cannot be taken literally* 
(though later, on p. 274, he himself offers a tentative solution), or with others (Prantl, Luria, 
Friedlander) that P. had not bothered to think out the consequences because he was only playing 
or joking ('Spielerei\ ' halb-scheriend\ 'playfully'; see Luria in next note and Friedlander, PL 1, 
256). The triangles recur at 81 b-c in a passage describing the physiology of youth, old age and 
death (p. 314 below). 
3 It was also the view of Eva Sachs, and in ancient times of Proclus, Simplicius and Philoponus. 
Aristotle, as we have seen, interpreted them as ideal or purely mathematical surfaces, as in modern 
times have Zeller, Archer-Hind and E. Frank. See Luria, Infinitesimaltheorie 151 with n. 120. 
288 
Construction of the primary bodies 
elemental fire may appear as burning flame, glowing embers or light, 
bright without heat, air as the limpid aither or as fog, and water as 
liquid or equally as solid but fusible metals or ice.1 He does not say 
explicitly that there was a strict mathematical relationship between the 
different sizes of triangle, but it was made clear at the outset and 
emphatically repeated in the summing-up at 69 b that the Demiurge 
introduced proportion and conmensurability (?????????) everywhere 
and in every way possible. Moreover if there were no such relation 
between the triangles forming varieties within the same element, one 
would have the curious situation that there could be no transformation 
between them, as there is between one element and another, though 
they resemble each other much more closely. Cornford's solution, 
whether correct or not, was beautifully simple, namely that in Plato's 
mind the triangles constituting the larger-grade solids are exact  
multiples of the smaller. Then the triangle of the smallest grade will be the 
highest common measure of the others, and will be the stoicheion 
proper, the irreducible element out of which they are built up. In  
mentioning six, Plato would be describing figures of an intermediate size to 
make it more immediately clear that the triangles can be put together in 
various ways {PC 234). The simplicity and rationality of this solution 
are attractive.2 
1 45 b, 58c-d. True to P.'s premises, melting and solidification of metals are not to be thought 
of as changes of'water* into 'earth* (in Vlastos's phrase, fusible metals are 'liquids with very high 
freezing-points', P.'s Universe 84), nor even as mixtures of the two, though mixtures of the 
elements do occur, e.g. the warming property of wine is attributed to an admixture of fire with 
water F0 a). They are due solely to differences of size between the octahedral water-particles, 
making them more or less mobile. Heat, the agent of melting, works by the action of fire- 
pyramids causing the preliminary disintegration of the icosahedra. (Details at 586-59a; cf. 
61 a 5-6.) 
2 Crombie finds it convincing (EPD 11, 220), but Popper has rejected it, and it has been 
criticized in detail by Pohle in I sis 1971. (See however Vlastos, P.*s Universe 69. V. gives an 
excellent summary of the whole theory.) One argument of Cornford's I do find surprising, 
namely that if the simpler procedure were followed the particles would have increased so rapidly 
in size that they might cross the threshold of visibility. (See his diagrams on pp. 237 and 238.) 
One assumes that the ultimate elements of body are microscopic, and we can surely imagine 
them as small as we like. Popper himself is inclined (with due caution) to attribute the subdivisions 
of the square and polygonal surfaces into 4 and 6 triangles, like the original division into 2 and 4, 
to Plato's interest in irrationals, and more specifically to the use of <Ji and ^3 to achieve an 
approximate squaring of the circle. See OS 1, 250-3. 
20 
289 
GHG 
Timaeus and Critias 
Perpetual motion and warfare of the primary bodies E7c 1-6, d 7-5833). 
By the axiom that like attracts like,1 the main masses of fire, air, water 
and earth collect in different regions, and would become wholly  
separate, and the universe completely static, were it not for the continual 
warfare between their particles whereby they are broken up and  
reunited in other shapes. Thus when sharp fire-pyramids attack an 
octahedral particle of air, and it re-forms into two fire-particles itself, it 
leaves the predominantly airy region to join the main mass of fire.2 It is 
a condition of motion that heterogeneous bodies should be in contact 
in the same area, one to cause the motion and the other to be moved; 
but given the fact of transmutation, even if they began like this, what is 
to prevent them from becoming separated and quiescent in the end? The 
cosmos, with all its movement and change, is to last for ever, and to 
ensure this Plato reminds us that his cosmology denies another main 
tenet of the earlier atomism, namely that the atoms had infinite space to 
move in. Plato's cosmos is a revolving, finite sphere,' the shape which 
contains all other shapes' C3b). From this envelope the particles 
cannot escape, but are turned back on themselves, jostling and thrusting 
at each other unceasingly. The smaller penetrate the interstices between 
the larger3 and proceed to break them up, while elsewhere the larger 
force the smaller to combine, and so the process of transmutation and 
consequent local displacement swings this way and that for ever.4 One 
must remember the continuity between the pre-cosmic motions and 
the same motions as ordered by the Demiurge when he took them 
over. (See 53a-b.) Much of the description of the earlier state at 52d- 
53a still applies, for the Demiurge turned to his own use, so far as 
1 Especially prominent in the atomism of Democritus (vol. 11, 429 f.), whom beyond reasonable 
doubt P. has in mind. What he is describing now is the work of Necessity. 
2 Plato does not expressly state that the four main masses form concentric spheres, with fire 
on the outside and earth at the centre, but it would be taken for granted. (See Cornford, PC 246.) 
It must, one would think, form the background for the explanation of the popular terms 'heavy* 
and Might* at 62C-63C 
3 P.'s denial of empty space within the cosmos at 5837 is not therefore true in the strictest 
sense: regular polyhedra in contact must always leave interstices. But none is surrounded by 
emptiness; contact is never lost. At 80c the denial of void is repeated and motion accounted for 
by the 'thrusting round* of the particles by each other. See further vol. 11, 147 with n. 1. 
4 One must assume that beyond the cosmic sphere there is not even space. This was the belief, 
not only of the pioneer Parmenides, but also of Aristotle, who added in astonishingly Platonic 
language and in line with the Phaedrus myth, that whatever is there is ageless and changeless, 
beyond place and time, in fact divine. See Cael. 279311-33. 
29O 
Construction of the primary bodies 
compatible with his purpose, the co-causes already furnished Of 
necessity*. 
Motion demands both mover and moved. In the course of making this 
point, Plato says E7e): 'It is difficult, or rather impossible, for there 
to be something that will be moved without something that will move 
it, and vice versa. In the absence of these there is no motion, and they 
cannot be on the same level.'1 Here he lays down a law that for  
movement to take place there must be two things, one which moves and 
another which is moved. Nothing single and homogeneous can move 
itself. But was it not Aristotle who, for reasons connected with his 
distinction between potentiality and actuality, argued, against Plato, 
that a self-mover was impossible? Plato's new principle, if applied 
universally, would seem to contradict his definition of soul or life in 
Phaedrus and Laws as the first cause of motion by virtue of moving 
itself. All physical, mechanical motion, as of one billiard ball when 
struck by another, can be traced back to soul (in that case the player's 
intention), for only what lives can initiate its own motion and transmit 
motion to others. Here however the contradiction is only apparent, 
because the only motion in question is physical, mechanical motion. We 
are taken no further than the secondary causes whose author is  
Necessity, learning 'in what manner it is of the nature of the errant cause to 
produce motion' D8a). Plato does not go right back to the first cause 
of cosmic motion,2 the gift to the whole universe of a self-moving and 
rational soul. When dealing with secondary causes he borrows freely 
from the earlier mechanistic cosmologies,3 e.g. the collisions and blows 
1 ?????, not quite 'homogeneous* (Cornford, as if ?????) nor 'in equilibrium' (Lee, as if 
????????) but 'evenly matched*. The comparison is of power or strength, as in a tug-of-war. 
When motion and change occur, two parties have fought and one is beaten; it is a weaker fighting 
with a stronger (i6e4 and 57a6). In the cosmos the stronger may be either a sharp and agile 
particle attacking a blunter, clumsier one, or a large force of bigger particles surrounding and 
crushing a few smaller ones. Homogeneity is of course ruled out a fortiori The homogeneous and 
therefore motionless One of Parmenides still haunts cosmology. Cf. vol. n, 36. 
2 I say 'cosmic motion' because I have argued, against others, that the irregular pre-cosmic 
motion in the Receptacle, caused ultimately, as motion in the cosmos is proximately, by  
heterogeneity and disequilibrium, has nothing to do with soul, either rational or irrational. 
3 At 58b4 he uses for 'compression* the word ???????, a technical term of felting which 
according to our sources had been used metaphorically by cosmologists from the early Milesians 
onwards. See vol. 1, 90 (Anaximander), 121 with n. 3 and 133 (Anaximenes), 391 (Xenophanes), 
and for other Presocratics the word-index to DK. 
291 
20-2 
Timaeus and Critias 
of the atoms which Democritus had identified with Necessity (vol. ?, 
404). His criticism is directed not so much against their description of the 
processes of nature as against their confusion of processes with causes. 
What escaped them was the ascendance of Reason over Necessity. 
Five worlds? At 3ia-b (pp°. 275 f. above) Plato gave his reasons for 
believing the cosmos to be unique. Immediately after the description 
of the five regular polyhedra at 5 5 c, he reiterates this and dismisses with 
contempt the idea of an indefinite number of worlds, but adds that 
someone might reasonably ask whether there were five. Nobody knows 
why. The conjecture of ancient commentators (for whom see Cornford, 
PC 220 f.) that kosmoi refers to regions within our world seems ruled 
out by the context, which, however, strongly suggests a connexion with 
the five figures just enumerated. Yet even if the dodecahedron stands 
for a fifth element, there seems no reason to conjure up five worlds 
each composed of one element. Miss Nicol suggested to Cornford (see 
PC 221 n. 3) that since only four of the five solids are assigned to 
elements, leaving out the dodecahedron, five worlds could be obtained 
by including it and omitting each of the others in turn. Some member 
of the Academy might have suggested that there was nothing to prevent 
the Demiurge from creating elsewhere a cosmos in the shape of a 
pyramid, cube, octahedron or icosahedron (though Plato would have 
been quick with his objections). But would these shapes be supposed to 
retain their connexion with a bodily element? At this point one finds 
oneself beginning to wonder what it would be like to live in a world 
with an outer shell composed of earth, and it is time to dismiss the 
gentleman with his unspecified 'other considerations',1 as Plato does 
without argument. 
The soul of the cosmos C4b~36d) 
In taking the world's body before its soul, we have given in to Timaeus's 
* random' way of talking, for he is careful to point out that as senior 
partner the soul must have been created first.2 The heavenly bodies, on 
1 Possibly Speusippus (H. A. S. Tarrant in Phron. 1974, 132, 137). 
2 Since the Greek word ??????????? is ambiguous between seniority of age and of rank or 
status, he goes out of his way to emphasize that soul is senior to body in both respects, ??? 
??????? ??? ????? ??????? C4C4)· 
292 
Soul of the cosmos 
the other hand, as already noted, are brought in after the soul in 
explanation of time. 
As a created god, divine but embodied, the cosmos combines all 
psychic functions, self-locomotion (of the most perfect kind attainable), 
true doxai about the sensible, and full knowledge of the intelligible 
C7b-c). The account of its making is highly symbolic, and the key to 
understanding it lies in the old doctrine, so prominent in Empedocles 
(vol. ?, 228f., 256) and still upheld here, that like is known by like.1 
The Demiurge proceeds in two stages. 
(i) Preparation of ingredients. This is described in the most difficult 
and debated sentence in the whole dialogue C5ai-b3). However one 
reads and interprets it, further thought suggests unsolved difficulties. On 
the most probable interpretation, there are three entities, Being (?????), 
Sameness (? ?????? ?????) and Difference, each of which has two forms, 
'the indivisible and ever constant' and 'the divisible which comes to be in 
bodies';2 that is, the Form^ and its copies in the sensible world. From the 
indivisible and divisible forms of each of the three, the Demiurge made 
an intermediate blend, then mixed the three intermediates to make 
the stuff of the cosmic soul.4 The significance of this will emerge later. 
1 ??????????? ?? ?????? ?? ?????, as Aristotle phrases it with express reference to soul in 
Tim. (De an. 404 b 17). 
2 P. speaks here again of ????? ????????? C532-3), but as in the Philebusy without promoting 
????????? to the full status of unchanging ????. See p. 233 n. 3 above. 
3 It must be so, and Cornford wrote {PC 64): ' The being of a Form is indivisible. A Form 
may indeed be complex and hence definable. But it is not . . . "put together" out of parts that 
can be actually separated or dissolved.' But P. does speak of the species contained in a generic 
Form as its parts (???? 31 a6), and though the Form Man is thus a part of the Form Animal it 
has none the less a separate existence. As I have suggested before (pp. 150, 270), the magnificent 
conception of these divine, eternal realities, though never abandoned, does not always stand up 
well to the development of analytical methods in P.'s later period. (Form and particulars are 
again compared at 516-52 a.) 
4 So Proclus (whom Cornford followed, also Ross, De an.y p. 177, Jowett's edd. vol. in, 
669 n. 3) and before Proclus Aristides Quint. De mus. bk 3 (p. 125 Winnington-Ingram), which 
I quote for its clarity:' The divine Plato, too, says in the Timaeus that the Maker of the soul took 
a mean between the indivisible and the divisible Being and combined with the intermediate form 
of Being the intermediates between the divisible and indivisible forms of Sameness and Difference, 
making a blend of the three.' Taylor's explanation, which resembles Martin's, is simpler (Comm. 
109; cf. Martin 1, 346). He regarded Same and Different as equivalent to the two forms of Being, 
indivisible and divisible, and wrote: 'He first takes two ingredients A and 2?, and by blending 
them produces an intermediate C. He then makes a single uniform whole by blending Ay By and C.' 
On the other hand (a) In spite of Friedlander, PI. in, 366, it is strange to speak of three ingredients 
in a mixture (???? 35 a6, ?? ????? bi) if the so-called third is a blend of the first two; (J>) Later, 
at 37a, P. speaks in the plainest terms of 'the Same, the Different, and Being, these three'; 
(c) Being, Same and Different figure prominently in Soph, as separate Forms (p. 151 above). 
293 
Timaeus and Critias 
(ii) Construction C5b—36c!). This is steeped in Pythagoreanism,1 
and its use of material imagery to represent the invisible psyche is not a 
little fantastic. The compound is treated as a sort of dough, to be 
kneaded, cut into strips and bent into circles. (Later, at 41 d, there is a 
reference to the * mixing-bowl' in which it was blended.) As in body, so 
in soul the primary requirement is proportion and harmony. The 
Demiurge therefore cuts off2 seven portions of the soul-fabric—now, 
it would seem, imagined as a long strip—in proportion, beginning with 
1 (outside the number-series according to the Pythagoreans) and  
proceeding in a series of square and cube numbers thus: 1, 2, 3, 4, <?, 9, 27. 
He then inserts harmonic and arithmetical means between each term in 
the original series.3 Squares and cubes suggest the three dimensions of 
body, which the soul in its cognitive aspect must recognize, but the 
numbers have also a musical significance. The harmonic mean was so 
called because it expresses the numerical ratios between what were for 
the Greeks the principal musical intervals.4 The word harmonia  
combined so closely the ideas of numerical ratio and musical 'concord* 
(????????) that they could not be separated in Plato's mind. But the 
arbitrary range of the scale presented by his table—four octaves and a 
major sixth, much greater than any employed in the music of the day 
(Taylor 140)—led Cornford to suppose that the compass was solely 
due to Plato's wish to end the series with 27, the cube of 3, and 'this 
decision has nothing whatever to do with the theory of musical  
harmony ' {PC 67). This goes too far. The soul is about to be distributed 
to move the stars and planets in their respective orbits, and as the 
Republic has it E3od): 'As our eyes are made for astronomy, so are our 
1 For the Pythagorean origin of P.'s amalgam of mathematics and music, and its cosmological 
significance, see vol. i, ch. iv d, esp. pp. 206-14, 220-4. 
2 The procedure would be easier to visualize if he simply, as Taylor says, 'marked ofF 
divisions as on a ruler or tape-measure. But P.'s word is ?????????? C632). 
3 The full scheme is set out by Cornford on p. 71 of PC. For the complex mathematical details 
of the passage see also Taylor, who quotes extensively from the ancient commentators, and 
Rivaud, 77m. 43-52. A. Ahlvers, Z. und K. bei P.y offered a new interpretation of the division 
of the world-soul, on which see Trevaskis, CR 1957, 31. 
4 The harmonic mean was defined by Archytas in his De musica (fr. 2 DK): it occurs when 
' by whatever part of itself the first exceeds the second, the second exceeds the third by the same 
part of the third*. An example is 6, 8 and 12: A2 — 8)/12 = (8 —6I6. For further explanation and 
the connexion with musical notes see Taylor, Comm. 95 or Freeman, Pre-Soc. Phils. 115. 
Aristotle, though no Pythagorean, still defined ???????? as 'a commensurate numerical ratio in 
the sphere of high and low* {An. Post. 90319). 
294 
Soul of the cosmos 
ears for the movements of harmony, and these sciences are sisters, as the 
Pythagoreans say and we agree/ The 'harmony of the spheres', 
poetically described in the Republic* was in Plato's mind even if not 
expressly mentioned. 
The essential is that the world's soul, being good and wise and 
destined to be everlasting, displays inner harmony,2 or due proportion 
and measure, which we know to be a prerequisite of goodness. It 
'partakes of reason and harmony' together C6c). Having now used up 
the whole of his mixture C6b 5-6), the Demiurge continues his self- 
imposed task by cutting it into two strips, laying one across the other in 
the form of a chi (X), and bends them round to form two circles, an 
inner and an outer, lying obliquely to one another. The immediate 
references are now astronomical, but for Plato the soul of the cosmos 
retains its twin powers throughout, the merely motive and the rational. 
The circles, being each a portion of rational soul, are in rotatory  
motion^ which they will impart to the heavenly bodies when these are 
created. 
The astronomical details are highly condensed, indeed elliptical, and 
Timaeus himself says later that to understand them properly would 
require a visible model Doe-d).4 In outline the scheme is this.5 There is 
1 Rep. 617b. For the harmony of the spheres see vol. I, 295. 
2 Perhaps a reminder is advisable that this is not the doctrine of soul as a harmony which was 
refuted in Phaedoy namely as supervenient on a harmonious disposition of the bodily parts. Soul 
is a harmony of its own parts. See vol. 1, 307-17. Its priority to body could hardly be more strongly 
expressed than by saying, as here, that it was created before it. 
3 The strange conception of a soul performing locomotion has to be accepted. Soul is not 
simply the power of a living creature to move itself: soul itself'moves, and imparts its own motion 
to the complex of soul and body. ' Only what moves itself is the source and cause of motion in 
others* {Phdr. 245 c, cf. Laws 896a). This evidently did not for P. carry any connotations of 
existence on a physical plane, like the air-soul of earlier thought, which he consistently denies. 
Sometimes, as at Laws 8966-973, he speaks as if it initiated only the higher 'movements* of 
thought and emotion, but it includes every stage of animation. Aristotle criticized P. here for 
making the soul behave like an extended body (De an. 406b26ff.). Skemp (TMPLD 83) tries to 
defend P., but has to admit on the next page that certain features 'do indicate localisations of 
???? *. They are especially apparent in the distribution of soul through the human body F9dff.). 
See also pp. 315-17 below. 
4 For the use of armillary spheres by P. and his contemporaries see Cornford, PC 74-6. It 
could explain the talk of rings rather than spheres. In any case, P. did not of course hold Aristotle's 
theory of the heavenly bodies as carried round in spheres of a tenuous matter (aither). 
5 For details and difficulties see Taylor and Cornford G2 ff.), and Dicks*s criticisms of Cornford 
in his Early Gk Astron. i24ff. An excellent account of P.*s astronomy, of the state of the science 
in contemporary Greece, and the effect of his metaphysical approach on its future development, 
is now available in Vlastos*s P*s Universe, ch. 2. 
295 
Timaeus and Critias 
(i) a rotation of the outermost sphere, the motion of the fixed stars 
from east to west in the plane of the celestial equator, accomplished in 
24 hours. It is called the motion of the Same, presumably (in this  
connexion) because it is also imparted, as the same motion, to the whole 
contents of the universe.1 It is thus the ' dominating' motion (has 
??????, 36c7). B) Contrasted with this is a revolution from west to 
east, and in the plane of the ecliptic, that of the Different, conceived 
both as a whole and as split into seven separate circles, those of sun, 
moon and the five known planets. All these bodies will share one general 
motion (besides, of course, its reverse, the motion of the Same), but 
some will also perform their own, differing from one another in speed 
and even direction. These proper motions counteract the common 
revolution of all, and explain why the planets do not all complete their 
orbits in the same time, the sun in one year, Jupiter in about 12 and so 
on. Sun, Venus and Mercury revolve only with the combination of the 
movements of the Same and the Different; Moon's extra motion is in 
the same direction as that of the Different but much faster; Jupiter, 
Mars and Saturn have extra motions in the opposite direction to the 
common revolution of the Different, which slow down, in varying 
degrees, their apparent motion round the circle of the Different. The 
whole scheme depends of course on the assumption, common to Plato, 
Eudoxus and Aristotle, that the apparently irregular paths of sun, moon 
and planets are reducible to a combination of perfectly circular  
revolutions.2 
1 Including perhaps the earth, but that is a notorious crux. See p. 306 ?. ? below. 
2 In a well known passage (Cael. p. 488 Heiberg) Simplicius says that P. himself set astronomers 
the problem of determining what uniform and regular (circular) motions must be presupposed to 
account for the apparent movements of the planets, and that Eudoxus was the first to solve it. 
(This appears, probably correctly, as a fr. of Eudemus in SpengeFs collection, no. 96; but Grote, 
PL 1, i24f., doubted the attribution to P., which he thought an embroidery by Sosigenes. For a 
full discussion of authorities and the historicity of the story, see now Vlastos, P.'s Universe, 
App. § l.) The faith in the perfection and circularity of the celestial motions haunted astronomy 
for the next 2,000 years. Even the young Kepler, with his deeply theological outlook, wrote that 
'we chose the spherical surface precisely because it was the most perfect quantity*. His Mysterium 
Cosmographicumy published in 1597 when he was 26, is a defence of the Copernican system, and 
its second chapter (quoted in full by Heisenberg, PCN 78-83) both mentions and echoes the 
Timaeus. This persistent exaltation of circularity has been commonly regarded simply as,  
scientifically speaking, a nuisance, but has been put in a more favourable light by Dicks {CR 1969, 362, 
mostly repeated in his Early Gk Astron. 176): 'Particularly in astronomy, the Greeks wisely 
placed more reliance on their advanced mathematical techniques than on observational data 
obtained by necessarily crude instruments . . . Granted that circular motion for the celestial 
296 
Soul of the cosmos 
Plato next mentions, in one sentence C6d), the contriving of the 
body of the cosmos within its soul, which envelopes it from centre to 
circumference, before passing to the cognitive aspect of the soul, 
devolving on itself, it made a divine beginning of unending and 
rational life' C6c). Its reason is to be imparted to the stellar gods and 
man, and life to the lower creatures as well, though in less pure forms 
as it encounters grosser and more perishable forms of body. We now 
learn the effect of its construction and motions on its cognitive powers, 
in a passage C3 a-c) which expounds what to us is perhaps the strangest 
feature of Plato's psychology, the association of circular motion with 
thought.1 
Since then it is blended of the nature of the Same and the Different and of 
Being, these three, and is portioned out and bound together according to due 
proportions and turns round upon itself, therefore whenever it comes into 
contact with either something that has a dispersed existence or something 
that is undivided, being moved throughout itself it tells what something is 
identical with and from what it differs, and in what particular respect, in what 
way, how and when it happens that things are severally related to and acted 
upon by each other with reference both to what becomes and to what is 
always the same. When true and consistent discourse, alike about what is 
different and what is the same, is carried on without utterance or sound2 
within the self-moved, and the circle of the Different, running true,  
communicates it to its soul,3 there arise firm and true judgements and beliefs; but when 
it concerns the intelligible, and the circle of the Same on its even course^ 
declares it, the result must be intellectual apprehension and knowledge. 
Besides exemplifying the Mike-known-by-like' principle, this passage 
brings into new company the logical distinctions established in the 
bodies did become accepted philosophical doctrine, yet this was in the first place a wholly  
legitimate inference from the results of observation, since the stars are seen to move in circular orbits 
across the sky, and sun and moon do appear to go round the earth in circles. What is commonly 
overlooked is the mathematical utility of the concept of circular motion; reduce your observed 
periodic movements to circles and combinations of circles, and at once you make them amenable 
to calculation and predictable as to both spatial position and time/ Cf. Taylor, Comm. 102 (on 
33*>7)· 
1 Here it is more or less assumed, but a demonstration is attempted in the Laws 897 dff. It goes 
back to Alcmaeon, for a comparison with whom see vol. 1, 351-7. 
3 Cf. the definition of thought in Soph. B63?) as 'dialogue taking place in the soul with itself, 
without utterance* (and similarly Tht. 189c). 
3 I.e. the soul of the self-moved regarded as the whole cosmos, the ?????? $???. 
4 I take ????? ??? and ???????? ?? at b7 and c 2 to be only literary variants. Cf. the hindrances 
to true judgements of identity and difference experienced by the newly-incarnated soul, before 
its circles have settled into their proper courses D4a, p. 310 below). 
297 
Timaeus and Critias 
Sophist.1 These had been introduced to refute the primitive, but at the 
time troublesome reasoning of Parmenides that one can say nothing 
about anything except that it ?, and what follows directly from that 
(its unity, continuity, changelessness). Plato showed that this limitation 
depended on confining the verb to one sense only, whereas in normal 
usage it had at least two: Exists' and 'is identical with\ Similarly 'is 
not' might signify either 'does not exist' or 'is different from'. In the 
light of this advance he named Being (or Existence), Sameness and 
Difference as the three universal categories. (See pp. 151-4 above.) 
One other point made in these meticulously drafted sentences is that 
the mind of the physical universe apprehends both what is and what 
becomes, the intelligible and the sensible, objects of knowledge and 
objects of belief. As elsewhere, a leading motif of the Timaeus is Plato's 
dualistic ontology and epistemology, expounded and emphasized at 
27d-28a and 5ib~52a. The mind's faculty is one of discrimination.2 
Whatever it encounters, it can say not only that it exists, but that it is 
identical with this, and different from that other. More precisely, it is 
capable of determining the relations (a) between one particular and 
another and (b) between a particular and a Form.3 We now see what 
was meant by making it a mixture of the indivisible and the divisible- 
by-bodies forms of its three ingredients C5 a). It is still the psyche of the 
Phaedo G9c-d), 'akin to' the eternal realities yet not itself one of them. 
It is 'between the worlds of being and becoming, at once intelligible and 
generated, indivisible and divisible, simple and in another way  
compound'.4 So constituted, it can apprehend both the intelligible and 
1 The relevance of Soph, is controversial. Cornford (PC61) asserted it emphatically, but it has 
been denied by Owen (SPM 327 f.) and queried by Manasse (Bucher uber P. 11, 83). Cornford 
perhaps went too far in saying that without Soph, the Tim. passage would be 'simply  
unintelligible*. It may be, as Crombie wrote (though such a generalization cannot be compelling) 
' unusual for P. to write in one dialogue words that can only be understood by the aid of a 
specific passage in another* (EPD 11, 262); but that the two are unconnected I cannot believe. 
Shorey was as emphatic as Cornford:' It is impossible to explain the world-soul as Rivaud does 
without mentioning the Sophist, and it is idle to affirm ... that the Same and Other of the Timaeus 
have nothing to do with the Same and Other of the Sophist* (CP 1928, 344f.). In Tht. A85 a-e) 
it is noted that judgements of existence, identity and difference belong to the mind, not the senses. 
2 In Greek ???????. Cf. Aristotle's description of sensation as ??????? ???????? ??????? 
at An. Post. 99b 35, and Crantor ap. Plut. An. procr. ioi2f.: the special function of the soul is 
?? ?????? ? ?? ?? ????? ??? ?? ???????. 
3 ???? ?? ????????? ?? ???? ??????? ?????? ... ??? ???? ?? ???? ????? ?????? ??? C7b2). 
4 Proclus, Tim. 11, 117 Diehl, quoted by Cornford, PC dy Cf. Plotinus, Enn. ??.2.1, ??????? 
??? ?? ?????? ?????? ?? ??????, ??????? ??? ???????? ?? ???? ???????. 
298 
Soul of the cosmos 
(through the medium of bodily sense-organs) the sensible, and can be 
dispersed in the bodies of all living creatures without losing its unity. 
Time and creation 
Nothing generated can be strictly eternal, that is, not simply everlasting 
but exempt from all distinctions of before and after, was and will be; 
but by introducing measure into the previous disorderly motions, the 
Demiurge effected 'a moving image of eternity, that which we call time 
(chronos), moving according to number', thus bringing the cosmos even 
closer to its pattern C7c-d). This distinction between the everlasting 
and the timeless may fairly be credited to Plato himself.1 Parmenides 
(whom Plato practically quotes) had said of his One Being: 'It was not, 
nor will be, since it now is, all together' (fr. 8.5-6). But for him there 
was nothing else, and about nothing nothing can be said. Plato rejects 
this outright denial of the world oidoxa, of temporal change or motion, 
and gives it a place, albeit a subordinate one, in a wider ontology. To 
this world chronos belongs. We translate it 'time', but the prevailing 
Greek conception of time was not ours. Chronos was cyclical and 
repetitive,2 and identified with the celestial motions which produce the 
recurrence of day and night, months and years, which Plato calls 'parts 
of time' C7 e 3); so for time to be, the stars and planets had to be 
created and placed in their orbits. The sun, made brilliant for the  
purpose, demonstrates in its daily revolution the motion of the Same, and 
by its own independent motion marks the year, as the moon the month. 
The circlings of the other planets are less easily observed, 'indeed men 
scarcely know that their wanderings are time' C9di), though there is 
in fact a' perfect' or great year, marked by the time taken by sun, moon 
and the rest to return to the same relative positions.3 Before the creation 
of these bodies, therefore, there could not be chronos, and to bring 
chronos into being was the motive for their creation.4 We speak of 
1 See however the full discussion of time and eternity in P. by W. von Leyden, PQ 1964. 
Owen has discussed the passage in Monist 1966, 332-6. 
3 Cf. vol. 1, 428-30 (where it is noted that the atomists were an exception) and Cornford, PC 
103 f.; also Arist. Phys. 223 b 23-33. This does not of course mean that Greeks generally held the 
Pythagorean doctrine of an exact repetition of history. The distinction between that and the 
cycle of celestial motions is clarified by Eudemus as 'numerical' and 'formal* identity respectively 
(fr. 88 Wehrli). 
3 On the Great Year in Plato and elsewhere see vol. 1, 282 and 458, and Cornford n6f. 
4 ??? ??????? ?????? 38c4. 
299 
Timaeus and Critias 
making clocks to measure time. For Plato chronos itself is a clock, not 
mere succession or duration but a standard by which duration can be 
measured. In Aristotle's concise definition (Phys. 219b 1) it is 'the 
number of motion in respect of before and after'. In the Timaeus the 
purpose of the sun is to provide' a conspicuous measure... in order that 
suitable living creatures [i.e. men] might possess number '. 
Suddenly we are reminded that the whole cosmogony is being seen 
in its relation to man, who through its providential arrangement can 
not only tell day from night and observe the seasons, but learn the art 
of counting.1 Through mathematics he can reach an understanding of 
the cosmic harmonia, and in this lies the secret of philosophy, by which 
the human soul itself is attuned to the divine music and achieves its chief 
end, 'assimilation to God as far as possible' (Tht. 176b; cf. Tim. 
90c-d). The starting-point is observation of the celestial motions, and 
the sense of sight is singled out for praise in a characteristic passage 
D7a-c): 
The sight of day and night, months and circling years, equinox and solstice, 
led to the contrivance of number, and gave us the idea of time and curiosity 
about universal nature, from which we have derived philosophy, the greatest 
gift of the gods that ever has or ever will come to mortal men . . . God 
invented and gave us vision in order that, by observing the circuits of 
intelligence (vous) in the sky we might use them to the benefit of the  
revolutions of our own thought (???????) which are akin to them, though disturbed 
while they are untroubled; and that by learning them thoroughly and being 
able to calculate them accurately according to their nature we might copy the 
unerring motions of the god [the cosmos] and give a firm basis to the errant 
motions in ourselves. 
Hearing too, through speech and music, contributes to the appreciation 
of harmony and helps to combat the discords in our souls. In  
fundamentals, Plato's philosophy changed little. The Timaeus only gives a fuller 
explanation of how, as we were told in the Gorgias, 'heaven and earth 
and gods and men' are united in society and orderliness, whence the 
whole universe got its name of kosmos, order; and in the Republic, 
'through association with the divine and orderly the philosopher 
1 Cf. Epin. 987 b-e. 
3OO 
Time and creation 
becomes divine and orderly as far as that is possible for a human 
being'.1 
It is surprising how many have taken the statement ' Chronos came 
into being together with the universe' ($6b6) as conclusive proof that 
Plato's story of the creation is metaphorical. 'No sane man', said 
Taylor, 'could be meant to be understood literally as maintaining at 
once that time and the world began together, and also that there was a 
state of things . . . before there was any world.'2 Space and becoming, 
Plato says E2dff.), existed 'even before the heavens came into being', 
and the contents of space, still untouched by the hand of God, were 
tossed hither and thither at random, in irregular and unbalanced  
movement without reason or measure. Without the heavenly bodies in their 
orbits, there cannot be chronos, but there can be what we must surely 
call time, that is, duration, a succession of before and after. 'Before and 
after', as Aristotle said (Phys. 223328), 'belong to motion, but chronos 
is these in so far as they are numerable.' It provides the regular and 
periodic units of motion by which duration can be measured. Space and 
time are not correlative to Plato. Space, the Receptacle of Becoming, 
was always there as the matrix on which the Demiurge set the stamp of 
order, but time is a part of the divine creation itself, a feature oikosmos. 
No one has put it better than Plutarch (Qu. PL 1007c): 'So Plato said 
that time came into being with the world (???????), but motion even 
before the world's birth. There was then no time, for neither was there 
arrangement, measure or mark of division, only an indefinite motion, 
as it were the unformed, unwrought matter (???) of time.'3 
1 Gorg. 508 a (vol. iv, 300f.), Rep. 500c-d. I make no apology for repeatedly drawing attention 
to this key passage (vol. i, 210, vol. iv, 500, 524^, and p. 246 above). 
? Comm. 69. Cf. 67: ' That he did not mean to say that there was ever a time when the world 
did not exist is plain from the express words of 38 b6 ?????? ???' ??????? ???????.' My point 
here was made by Hackforth (CQ 1959, 21 f.), who refers to Skemp, TMPLD 77. But what 
Skemp says is: 'It may not be a literal description of what happened in the past, but it is a 
description of the actual process of the world in which we live', which sounds rather like the 
'logical analysis' that he has just denied it to be. The removal of this particular argument does 
not of course settle the question whether the creation is intended mythically or literally; but many 
have thought that the mythical interpretation needed no other support. So e.g. Gauss, Handk. 
in, 2, 170. 
3 In any case, is the idea of a 'time before time' so inconceivable? Not to a nineteenth-century 
cosmogonist, C. S. Pierce, who wrote of'the first stages of the development, before time existed*, 
' Out of the womb of indeterminacy [the Receptacle as mother?] we must say that there would 
have come something, by the principle of Firstness, which we may call a flash. Then by the 
principle of habit there would have been a second flash. Though time would not yet have been, this 
3OI 
Timaeus and Critias 
The question whether Plato believed the creation of the cosmos to 
have been an actual event, or simply wished to convey allegorically the 
idea that it depends for its goodness and order on the divine will, is 
endlessly debated and may be insoluble, though earlier sections of this 
chapter have hinted at a preference. Belief in a creation does not of 
course commit one to obviously symbolic details1 like the mixing of 
Being, Same and Other in a bowl. On the other hand antipathy to the 
mythical interpretation could be in part an unjustifiable reaction from 
the unacceptable way in which the denial of a 'time before time' has 
been used by so many to prop it up. One can at least say something of 
the history of the debate and explain the alternatives further.2 
The argument starts among Plato's immediate followers. Aristotle, 
who believed that everything generated must some time perish, took 
the creation literally, and criticized Plato for saying in the Timaeus that 
the world has been generated but will last for ever.3 Elsewhere he says 
that Plato is the only philosopher to maintain that time had a beginning: 
'for he says that it came into being together with the world, and that 
the world was generated' (Phys. 251 b 17). He also mentions the  
alternative view {Cael. 279 b 33-28032): 
The self-defence attempted by some of those who hold that it is indestructible 
but generated, is untrue. They claim that what they say about the generation 
of the world is analogous to the diagrams drawn by mathematicians: their 
exposition does not mean that the world ever was generated, but is for 
second flash was in some sense after the first, because resulting from it.' (Quoted by Gallie in 
Pierce and Pragmatism n8f., italics mine.) 
1 As Hackforth remarked in an article full of good sense {CQ 1959, 20). 
? The most recent modern champions on either side are Vlastos, * Creation in the Timaeus: is it 
a Fiction?' (printed with his earlier article, 'The Disorderly Motion in the Timaeus*, in SPM), 
and Taran, 'The Creation myth in Plato's Timaeus* (in Anton and Kustas, Essays), which is 
expressly directed against the literal interpretation of Vlastos. This with G. E. R. Lloyd's  
discussion in P. and ?., 279 ff., will provide orientation in the modern controversy, though I would also 
single out Hackforth, 'P.'s Cosmogony', in CQ 1959. (H. changed his mind: contrast his 'P.'s 
Theism', SPM 442.) Cornford (like Taylor) took the creation as mythical, while admitting that, 
since this left the Demiurge without a function, he must be mythical too. He therefore identified 
him with the reason in the world-soul {PC 97); but P. was no pantheist, and no yiy???????, not 
even ? ????????* ??? ????????? B9a 5), can be its own cause. (Cf. 2834-6, as well as the Philebus.) 
3 Cael. 280a, 28-32. As Martin points out in an interesting note {Tim. 11, 195), this criticism is 
met in advance by Tim. 41 a, though Arist. would not agree. The world is by its own nature 
perishable, but preserved by the will of the supreme being who produced it (p. 279 above). 
This is what makes it an exception to the rule laid down by P. himself in Rep. E46a), ???????? 
????? ????? ?????. 
302 
Time and creation 
instructional purposes, since it makes things easier to understand just as the 
diagram does for those who see it in process of construction. 
For these men Plato's account of creation is no more than an analysis of 
the world's structure expressed in synthetic or genetic terms. Similarly 
a geometrician to describe the form of a cube may speak in terms of a 
square being constructed out of four equal straight lines, and then a 
cube out of six squares. He does not mean that lines exist prior in time 
to planes, or planes to solids, but has described the cube as if in the 
making, as a teaching device. On this view 'generated' means only 
'derived from an external cause, not self-born nor self-substantial' 
(Procl. Tim. ii, 276 Diehl). It goes back to Speusippus, Xenocrates and 
Crantor,1 and seems to have prevailed in the later Academy as among 
the Neoplatonists.2 
Modern opponents of the literal interpretation of creation express 
themselves similarly, e.g. Hackforth:3 Plato does not mean that either 
soul or the universe was created in time: 'The meaning in both cases is 
that they are derivative existents, things whose being depends on  
something more ultimate.' The Greek words for 'becoming' and 'to 
become' (???????, ?????????) had two senses: (a) coming into 
existence at a particular time, either suddenly or at the end of a process 
of development or manufacture; (b) in process of change, in which 
1 Simpl. Cael. 303 Heiberg (and cf. Vlastos, SPM 383 n. 1), Procl. 77m. 1, 277 Diehl. For 
Speusippus see fr. 54a and b and p. 31 Lang. Theophrastus was doubtful, but followed Arist. in 
maintaining that if P. did intend the world's construction only in the geometer's sense, the 
analogy was a false one. (Fr. 29 Wimmer, cf. Arist. Cael. 280a2-10.) (Here the comments of 
Taylor, Comm. 69 n., and Taran, I.e. 390, are a little tendentious.) We have not Xenocrates's 
words, but the Theophr. passage tells slightly against the suggestion of Vlastos and Hackforth 
that he and others might have put forward their view not as P.'s meaning but their own. It was 
their own (see e.g. Plut. An. procr. ioi3a-b), and Aristotle called it a selj-deience (??????? 
???????), but it is clear that, not wishing to go against Plato, they assumed that it was his too. 
This is what made Xenocrates, in the current phrase, 'overreact' (?????????????????*, ps.-Alex. 
in Ar. Metaph. p. 819 Heib.). See Lang, Speusippus 30. 
2 Simpl. (I.e.) attributes it to 'Xenocrates and the Platonists'. Augustine too (Civ. Dei 10.31) 
says the latter believed that beginning means causal subordination, not an order in time, and it was 
the view of the middle Platonist Albinus (Isag. ch. 14; see Gauss, Handk. m, 2, 189). But the 
opposite view was held by Plutarch, by Albinus's fellow-Platonist Atticus, and according to 
Proclus A, 276) by 'many other Platonists' (not therefore 'ein^ig Plutarch und Attikos', as 
Dorrie in RE, 2. Reihe, xviii. Halbb. 1523). For Plato himself it may be relevant to notice that in 
the Republic E27a) he ridiculed the geometers' habit of speaking in terms of action, of' squaring', 
' applying', ' adding' and so on, ' as if they were doing something'. (I have not seen M. Baltes, 
Die Weltentstehung des plat. Tim. nach den ant. Interpreter Teil I, 1976.) 
3 SPM 442, i.e. before he changed his mind and wrote his 1959 article. 
303 
Timaeus and Critias 
though something new is always appearing, something old passing 
away, the process may be thought of as going on perpetually. It does 
not then need a cause to start the process at one moment and complete 
it at another, but a sustaining cause, to keep it going endlessly.1 The 
latter sense had a peculiar importance for Plato, whose talk of'what is' 
and 'what becomes' marked a difference of ontological rather than 
temporal status. It should be observed, however, that Plato is aware of 
the ambiguity and more than once goes out of his way to remove it in 
favour of the temporal sense. The 'timeless dependence' explanation is 
not in Plato: it begins with Xenocrates.2 Consider the way in which the 
question is first put and answered: 'Has it always been, having no 
beginning of its birth, or has it come to be, having started [aorist 
participle] from some beginning?'3 And the answer is an emphatic 
perfect tense: 'It has come into being.' The reason is that it has a 
physical body, and such things 'can and do come into being'.4 Another 
relevant instance has been referred to already (p. 292 n. 2):' senior both 
in birth and in excellence' at 34c removes any misapprehension that 
'senior' (???????????) might refer to status only. 
Many more passages could be quoted,* but the case for denying 
literal creation does not rest on a claim that Plato avoids the language 
of temporal sequence.6 That Plato speaks of God creating the cosmos 
1 In this explanation I have drawn freely on Cornford's phraseology {PC 24 f.). He goes on 
to say that for (?), which is his own interpretation of P.'s meaning here, 'both the images 
"father" and "maker" are inapprorpriate' (though P. uses both!), and we should rather think of 
'some ideal or end, constantly exercising a force of attraction, and perhaps of some impulse in the 
thing itself, constantly aspiring towards the ideal \ He does not add that this is pure Aristotelianism. 
It accurately describes the action of Aristotle's god ('the Unmoved Mover') on the natural 
world, which excluded divine providence and so was quite foreign to P. 
2 A man pleasantly characterized by Martin {Tim. 11, 195) as 'homme de bien, tres-studieux, 
mais de peu de g?nie\ He was relying on D.L. 4.6, Aelian, V.H. 14.9, Plut. De recta rat. aud. 
18, 47c 
3 28b6ff. Cornford translates ???? 'source' at b6 and 'beginning' in the next line. It could 
mean either, but not both in this one short sentence. In defence of his view, C. (pp. 25 f.)  
emphasizes ??? in the phrase ?????????? ??? at 28 a, without noting that the authority for ??? in our 
printed texts is slender. See Hackforth, CQ 1959, 19 and Whittaker, Phoenix 1969, 181-5. 
4 ????????? ??? ??????? ?????. Cornford translates 'become and can be generated', but I do 
not believe, that the second epithet is merely pleonastic. 
5 E.g. ??? ???? ???????? ???? at 34a-b, which Proclus A1, 100 Diehl) does his Neoplatonic 
best to explain away, and the frequent use of aorist and perfect tenses: ????????? 28c, 31b, 
??????????? 29a, 39?, 31b; the stars ???????? 39d. ??? and ??(? are used at 52d4 and 5337. 
Cf. also ???????? ???? ??????? of air, 56 b. 
Yet how easy it would have been to omit the words ??? ??????? at 34c! Nor does the phrase 
??????? ??????? suggest myth. It is rather a technical philosophical expression for distinguishing 
304 
Time and creation 
by stages out of a pre-existing disorder is certain, but he could still have 
deliberately chosen to cast an analysis of nature into a synthetic or 
narrative form. As to that, I can only refer to what has already been 
said about tht intention of the 'probable mythos (or logos)' (pp. 250- 
3). In any case, the statement that the cosmos 'has come into being' 
belongs to the fundamental principles laid down before the 'probable 
story' is begun. But at this point the reader may be left to study the 
arguments and counter-arguments of Vlastos and Taran (and Cherniss 
to whom both refer frequently) and make up his own mind. 
Note. To compare the Timaeus with passages in other dialogues may 
savour of the use of biblical proof-texts, but one or two may be  
mentioned. If Pol. 273b ('The universe was in great disorder before it came 
to its present order') is disallowed as itself part of a myth, Laws 892a 
speaks of soul in the same temporal terms as Tim. 34c: it is older than 
body because it and its kindred 'came into being before bodily things'.1 
H. J. Easterling in Eranos 1967 maintained that Timaeus and Laws were 
consistent on the subject of causation, but has been criticized by Taran 
{I.e. 403). Taran however argues from the assumption that soul in the 
Timaeus is the ultimate cause of the pre-cosmic disorderly motion of 
the Receptacle. 
Creation of living creatures: nature and fate of the human soul C9e~42c) 
We might say that Plato's observation of four main types of living 
creature led him to assume the existence of as many archetypes. His 
argument is the reverse. Because the Model contained these types, the 
cosmos, to be complete, must contain them too. They are the race of 
gods (that is, principally, the stellar gods), the winged and airborne, 
aquatic and land animals. Each corresponds to one of the primary 
bodies. The divine race the Demiurge made mostly2 of fire, for beauty 
and conspicuous visibility, and set them in the circuit of the supreme 
(that is, of the Same) to adorn it all round. (But Plato says 'set in the 
intelligence of the supreme' D035), so complete is his assimilation of 
the senses of ????????. With Aristotle indeed it become wholly technical, and the antithesis of 
??????? and ???? (or ?????) ???????? is one of his favourites, e.g. Metaph. 105034, ?? ?? 
??????? ?????? ?? ????? ??? ?? ????? ???????. 
1 See on this pp. 366 f. below. 
2 Not entirely. With ??? ???????? ????? cf. Epin. 981 d. 
21 305 ghg 
Timaeus and Critias 
circular movement to reason.) Besides their common motion in this 
circuit each star performs a revolution in its own place ' as each thinks 
the same thoughts about the same things'. Earth is at the centre, and 
described as our nurse and guardian, maker of night and day, and more 
surprisingly, 'first and eldest (or most venerable, ??????????) of all 
gods born within the heaven'.1 Timaeus then excuses himself from 
recounting the birth of the gods of ordinary Greek mythology. That is 
the province of the poets, who claim to be of divine descent and must 
know their own family history. This may be no more than half ironical 
—Plato was a staunch upholder of the established cults—but does 
show that traditional mythology is to play no integral part in the 
Timaeus. 
As created beings, the lesser gods cannot be immortal, but will live 
for ever in dependence on the will of their creator. It is for them to make 
and nurture the three mortal kinds, which if made by himself would be 
as gods.2 Only the divine and immortal part of the human soul will he 
make himself, then hand it over to them to 'weave mortal and immortal 
together', creating the mortal parts of the human soul and implanting 
the whole in bodies, it appears that, though birds, fish and animals will 
be needed to make the world 'complete', the gods will at first make only 
men. Women and lower animals are hastily dismissed in a postscript 
1 I shall not go into the vexed question of the possible motion of the earth. The different 
interpretations of the relevant passages in Tim. and Aristotle Dob-c, Cael. 293 b 30 and 296326) 
have now been fully discussed by Dicks in his Early Gk Astron. A32-7), who concludes (with 
Cherniss) that A. does not attribute to P. a belief in a moving earth. There I must leave the matter, 
though with an uneasy feeling that this involves an unnatural rendering of A.'s Greek. It may 
seem odd that in P.'s cosmos the earth—heavy, solid, furthest from the heaven, home of transient 
beings—should be ????? ??? ??????????? But one must remember the status of Gaia in Greek 
belief and cult. Sprung direct from Chaos and mother even of Ouranos (Hes. Th. 117, i26f.), 
* first of prophets' (Aesch. Eum. 2) and original mistress of the Delphic shrine, she was indeed 
our nurse (Aesch. Theb. 16) and the first and most venerable of deities, and is about to appear as 
such in the summary of popular theogony which immediately follows D0e). Plato will not show 
disrespect to deep-rooted Hellenic beliefs, though concerned, as he showed in Rep., to purge 
them of certain moral crudities. 
2 One might ask, 'And why not? Why must they be mortal?' Taylor says (Comm. 253) 
* Death and generation are part of the original good plan of God.' This is wrong. The Model 
does of course contain eternal Forms of what in the cosmos will be mortal creatures C0 c), but 
Archer-Hind was nearer the mark when he said that becoming and perishing belong to materiality 
{Tim. 140). God can only make from a given unformed material something as like as possible to 
the non-material Model, and the implanting of souls in bodies is one of the limitations imposed 
by Necessity D2 a). Even the cosmos and the star-gods have bodies and are generated and not 
immortal, though the Demiurge has power to postpone their dissolution indefinitely. 
306 
Creation of living creatures 
(9oe~92c) after a reminder by Timaeus that he only agreed to take the 
tale of creation as far as the birth of mankind. According to the 
'probable story' (90 e 8) the rest originated from souls of inferior men 
degraded by transmigration into lower forms of life: first women, then 
birds from harmless lightweights who thought it enough to study the 
heavens with the eyes alone. (Cf. Rep. 528e-30c.) Quadrupeds and 
reptiles come from men whose minds were subordinate to their animal 
desires, and fish from the stupidest of all, unworthy any longer to 
breathe the air and thrust into the turbid waters. Suitable bodies are 
devised by the lesser gods. 
I should not like to say just how far this appendix is a mere jeu 
d9esprit. The closest parallel is Phaedo 81 d-82b, but neither there nor 
elsewhere does Plato repeat this insult to women as originating from 
morally defective souls (90 c), which is scarcely compatible with their 
role in the Republic (though that has actually been recalled at the  
beginning, 18 c) or the Laws.1 In the Phaedo the only degenerate forms of life 
mentioned are bees, wasps and ants, donkeys, wolves and birds of 
prey. Moreover Timaeus's words at 42a, 'human nature being twofold, 
the superior part is the one later called "man" (????)', strongly 
suggest that both were there from the beginning. Perhaps the most 
likely explanation is that Plato does sometimes make Timaeus speak in 
his character as a Pythagorean, and for a Pythagorean the female came 
in the second column of opposites, along with evil, darkness and the 
unlimited.2 The rest accords with the Phaedrus and Republic myths, 
though without mention of the choice of lives. What one can say is that 
the ethical implications of the Orphic and Empedoclean doctrine of the 
cycle of births and the means of escape gave it a strong attraction for 
Plato. 
Making and destiny of human souls D1 d~42e). The Demiurge now made 
a new soul-mixture,' in the same bowl in which he had formerly mixed 
the soul of the whole', and using the remains of the same ingredients 
1 That women fall short of men in ????? {Laws 781c), and men are superior and stronger 
{Rep. 455 c, 456a), does not alter the fact that women are to be trained for and share all activities 
and duties with men, including government. 
2 For the table of opposites see vol. 1, 245 f. The above suggestion is made by D. F. Krell in 
Arion 1976, 401. (This does not imply that I agree with everything in his article.) 
307 
21-2 
Timaeus and Critias 
but less perfectly blended.1 Then dividing it into separate souls equal in 
number to the stars, he set each one on a star * as on chariots' (an echo 
of Phdr. 247 b), and lectured them on the laws of the universe and their 
own destinies. Their implanting in bodies was a matter of necessity 
D233-4), but he himself would ensure a fair and equal start for all. If 
they yielded to the passions and desires that incarnation would bring, 
they, not he, would be responsible for the consequences.2 Whoever 
followed the circuits of reason (that is, of the Same, 42C4~d2) would 
return to a blessed life on his appointed star. Once again Plato shows 
his skill in conveying the same truth in slightly different mythological 
form. In the Phaedrus B48 a ff.) only souls which have seen reality from 
the rim of the circling heaven are born as men, for only human souls 
can pass beyond sensations to concepts and if philosophically inclined 
and trained retrieve their ante-natal knowledge and regain their wings. 
The actual connexion with stars is borrowed, like so much else, from 
popular tradition. (Cf. Aristoph. Peace 832f.) 
What Plato is trying to describe in this imagery is an immortal soul 
(for the Demiurge himself is only concerned with the immortal part) 
which is yet in some way inferior to the cosmic soul because it is 
capable of being incarnated in mortal bodies. The bodies animated by 
the world-soul are those of the everlasting gods, sun, moon, stars and 
planets. The same contrast is pictured in the Phaedrus through the 
image of the charioteer and winged horses. This presents on the one 
hand the souls of the gods, a blessed throng who always remain on the 
outermost rim of the heavens; and on the other, those destined for 
incarnation in mortal bodies. The difference is symbolized by saying 
that in the souls of the gods all three constituents—driver and both 
horses—are alike good, whereas in the others only the driver  
(representing the philosophic eros) is perfectly good. The horses stand for the 
'spirited' and appetitive parts in Republic and Timaeus, and the bad and 
unruly one finally brings about the fall of the soul to earth and its 
incarnation in a mortal body. Through this image Plato expresses a 
1 Other translators say 'less pure', and Jowett even adds 'diluted'; but with what could a 
mixture of Being, Same and Other be adulterated? ?????-ros (from ???????, not ?????????) 
means unspoilt or unharmed, and is explained by ???? ????? ???????. They are not mixed in 
such exact proportions. 
2 42d3~4; as in Rep. F27c). 
308 
Creation of living creatures 
religious truth (the fall of a soul made in the image of God) which 
cannot be explained rationally.1 In the Timaeus he tries what is surely a 
less happy metaphor, of the Creator stirring a sort of pudding in a 
basin. Human souls are made of the same mixture as the souls of the 
gods, but its quality has somehow deteriorated. Once the fall has 
occurred, Plato's dualistic philosophy can account for the evil in 
human life, as especially in the Phaedo, by the contamination of soul by 
body, but it can never explain how the divine soul came to be incarnated 
in the first place. The passions are definitely the result of incarnation: 
they do not precede it. It happens 'of necessity' D2a), not therefore 
expressly through a fault or flaw in themselves, as in the Phaedrus and 
Empedocles,2 but through an external force or inevitability. At the same 
time Ananke does stand for the element of imperfection ('disorder' in 
Timaean terminology) in things, a certain intractability in their nature, 
so it is probably fair to connect this phrase with the imperfect harmonia 
in the material of souls destined to enter bodies. 
The supreme Creator now retired from his work D2 c), and the lesser 
gods turned to their appointed tasks. To make our bodies, they used 
portions of the four elements in the cosmos, as, for our souls, the 
Demiurge had used the remains of the cosmic soul-material. As in the 
Philebus (p. 215 above), the kinship between microcosm and  
macrocosm is emphasized at every turn. These portions are 'borrowed to be 
repaid' D2c), for mortal bodies are at death resolved back into the main 
masses of earth, water etc. from which they came. The details of our 
creation come much later in the narrative, at 6c)cfF. Taking over the 
immortal soul, the gods gave it a mortal body as a 'vehicle', and 'built 
on' another kind of soul, the mortal, containing the irrational feelings 
and emotions, headed by pleasure, 'the greatest lure of evil'. The  
complete soul of mortal man is, as in the Republic, tripartite, and each part 
1 This is elaborated in the section on Phdr., vol. iv, ch. vi C). 
2 The whole scheme of incarnation and transmigration is reminiscent of Empedocles, and the 
poetic picture at 43b-c of the hostility of the several elements to the soul vividly recalls fr. 115, 
which describes the sufferings of daimones exiled from the company of the blessed and born into 
various forms of mortal life. * The mighty air pursues them to the sea, the sea spews them out to 
the dry land, the land to the rays of the flashing sun, who casts them into the whirling air. One 
receives them from another, and all loathe them alike. Of these am I now one, an exile from the 
gods and a wanderer.' Behind both P. and Empedocles are the Orphic poems, if what I have said 
in ch. 11 of G. and G. is right. 
309 
Timaeus and Critias 
is assigned to a specific part of the body.1 The immortal part, the reason, 
resides in the head, separated from the others by the narrow channel of 
the neck, to avoid contamination. (This example may serve to illustrate 
the teleological character of all Plato's physiological descriptions.) Since 
the mortal part itself has a nobler and a baser half, another barrier is 
formed by the midriff. The ' spirited' part2 inhabits the thorax with the 
heart as its organ, near enough to the reason to heed its behests, and the 
appetites are below in the belly,' tethered like a wild beast, untamed but 
necessary to be fed if a mortal race were ever to exist'.3 
The infant soul and the cause of error D3a~44d). When first the gods 
confined an immortal soul within the flux of a mortal body, it produced 
a horrific effect, repeated at every birth today D437-9). Plato's figure 
throughout is of something at the mercy of a raging torrent, or the ebb 
and flow of a tide. It is the Heraclitean world-picture which, as Aristotle 
said, he never gave up. In this welter of instability the circuits of the 
soul, though indissoluble save by him who made them, have their 
original motions twisted and disorganized, just as the human body 
rushes in all directions in contrast to the uniform revolution of the sky. 
Lacking the regularity of the turnings of Same and Other in the cosmic 
soul, newly incarnated souls cannot produce rational thought, but are 
full of falsehood and folly. ' When they meet something outside to 
which the name Same or Different applies, they speak of it as the same 
as this and different from that contrary to the true facts.' Plato's view 
of children ('harder to handle than any wild animal', Laws 808d) is 
the reverse of the sentimental or Wordsworthian. Our birth is indeed a 
forgetting, but far from 'trailing clouds of glory' until 'shades of the 
prison-house' descend with advancing years, the shock of birth and the 
flood of undigested and unclassified sensations, which play a far greater 
1 This is in conformity with the ostensibly scientific spirit of this part of Tim., and leads to 
much psycho-physiological detail, e.g. the throbbing of the heart in fear and anger and the 
provision of lungs to relieve it G0 c). Such localization seems to threaten the complete immateriality 
of the soul, but P. would still distinguish between any form of consciousness and the physical 
organs through which it is experienced. Cf. Tht. 18^-85 e: the eye does not see; the psyche sees 
through the eyes as its instruments. Also pp. 314-17 below. 
2 The nature of this Fupos or ?? ????????*) has been described in vol. iv, 474. 
3 70c 'Necessary pleasures' played a part in Phil. F2e, p. 233 above). On the position of 
sexual desire see Additional Note (i) on next page. 
3IO 
Creation of living creatures 
part in childhood than later, overwhelm the mind and drive from it all 
that it has seen and learned in its unbodied state. With advancing years 
the cessation of physical growth and development allows it comparative 
calm in which to recover its regular courses and a rational life becomes 
possible. With proper training a man may become wholly sound, but if 
neglected he will return to Hades after a maimed life with his  
imperfections and folly upon him.1 
The problem of the possibility of error is a serious philosophical one, 
and as such Plato discussed it in the Theaetetus. Here it takes its place 
in the great myth of the creation of all things, and appears in symbolic 
guise. One has to remember that he thought it worth while to put the 
same problems in both a dialectical and a mythical setting, a feature 
which sets him apart from every other thinker. But in the Theaetetus and 
even the Sophist the question remained unanswered. Perhaps3 it could 
only be answered in the light of a vast scheme of theology and  
cosmology, and man's situation within it, in which he believed none the less 
profoundly because it could only be expressed in the form of a probable 
account or mythos. 
ADDITIONAL NOTES 
(/) The status of eros 
Of the appetites only those for food and drink are mentioned G0d 7), and 
Cornford (PC 292 f.) suggested that the omission of sexual desire from the 
lowest part of the soul is deliberate and due to the exalted position of eros in 
Plato's philosophy, especially in the Symposium. In the strange amalgam of 
spiritual and physical which marks the Timaeus, the semen is a part of the 
marrow and through the spinal cord continuous with the brain, lodged in the 
seat of reason. (So 73 c-d. This was a theory of the Sicilian medical school. 
Some Hippocratics believed the semen to come from all parts of the body, 
solid as well as liquid: see Cornford 295 and Hippocr. Degenit. 3, vn 474 L.) 
The lower organs only provide a receptacle and outlet for it. Aristotle also 
held that, as the producer of life, semen both has soul and is soul potentially 
(Gen. an. 75 5 a 7). I am doubtful of Cornford's point however because (a) sex, 
at least as a means of reproduction, is excluded because women are not 
1 As an example of how P.'s thought underwent development and substantiation rather than 
radical change, one may compare this with the Phaedos statement of our loss of knowledge at 
birth G5 e; vol. iv, 345). 
* Perhaps, because we do not know what the Philosopher would have contained (see pp. 154^ 
above); but the Philosopher was never written. 
311 
Timaeus and Critias 
supposed to be yet created, and (b) when sexual reproduction is introduced, 
the male sexual organ is described as 'disobedient and imperious, like an 
animal heedless of reason, bent on mastery through frantic desires' (91b). 
' The eros of reproduction' (91 b 4) could easily have been mentioned here as 
belonging to the lower part of the soul; and in fact the phrase at 70d, 'the 
part of the soul desirous of food and drink and everything of which through 
the body's nature it stands in need', is fairly inclusive. Cf. 69d: in making 
mortals die gods included 'eros that ventures all things'. 
(ii) Extra-terrestrial life? 
In his speech to the lesser gods, the Demiurge says at 41 c 8 that having sown 
the souls and made a beginning, he will hand them over for the creation of 
the mortal parts of soul as well as body. At e4, telling the newly-created souls 
their fate, he says that ' having been sown each into its proper instrument of 
time, they must be born as the most god-fearing of creatures'. After these 
two forecasts, the actual operation is described at 42d2: 
' Having given them all these ordinances that he might be guiltless of the 
future evil of each one, he sowed them, some on the earth, others on the 
moon, and others on the remaining instruments of time [the planets], and 
after the sowing he handed them over to the young gods to mould bodies, 
to create the other parts of a human soul that needed to be added, and to 
govern and guide the mortal living creature as well as lay in their power, 
save in so far as it might become a source of evil to itself.' 
I have put these passages together because it has become a matter of 
controversy whether the statement that souls are ' sown' not only on earth 
but also on the sun, moon and planets means that these as well as the earth 
are supposed to be inhabited. It is an exciting thought. Taylor believed it did 
(Comm. 258f.), Cornford denied it (PC 146 n. 2), Hackforth in his lectures 
said that though it seemed absurd to think of human beings as living on the 
sun, one might conceive of some kind of embodied souls living there. 
That men lived on the moon seems to have been a Pythagorean belief 
(DK 44a20; vol. 1, 285), which has also been attributed, on doubtful 
authority, to Anaxagoras (Guthrie, OGR 247 n. 10; HGP 11, 308 and 314). 
Of any similar belief about the sun or planets there is no trace anywhere in 
Greek thought. Nor should it be necessary to go into the question here, for 
the passages just quoted surely show that it is not what Plato meant to say. 
In all three, it is only after the Demiurge has sown the souls in the planets 
that they are to be provided with their lower elements and incarnated in 
bodies. Three stages are involved: (i) journey round the outer heaven on 
star-chariots to learn 'die nature of the whole', die knowledge which they 
312 
Creation of living creatures 
may recover by their own efforts in the coming bodily life; (ii) 'sowing* in 
the instruments of time: still only the immortal soul is concerned; (iii) 
addition of other parts of soul and setting of the whole in a body. If this is 
right, the intermediate stage may symbolize, as Cornford suggested {PC 146), 
the intermediate status of the soul, partaking of both being and becoming, 
subjected to time and change yet essentially immortal; but one can only 
conjecture. 
Necessity and design in the natures of men F1 c-c)od) 
Teleological explanation. The rest of the dialogue deals in detail with the 
limitations imposed by necessity on the nature of human bodies and 
faculties, and the devices of the gods to overcome them in the interests 
of the immortal part of the soul.1 Its subjects include anatomy and 
physiology, the composition of bone, hair, flesh and so on, the function 
of certain organs and of the blood, respiration and digestion, and also 
pathology, the diseases of both body and soul, their causes, prevention 
and treatment. Here we can only note a few general features and one or 
two particulars which have a wider interest. 
Teleological explanation is universal, and sometimes carried to 
lengths which have aroused suspicion that Plato is only amusing  
himself. Wilamowitz was moved to write {PL 1, 612): 'The description of 
the human body, its parts and their functions, would draw from an 
unprepared reader (if he had not thrown the book away at once) the 
cry: "This may be madness, but there's method in it." Sometimes he 
might feel doubtful even about the method.' As throughout the 
dialogue, Plato warns that no account of such matters can claim  
certainty G2 d), and we have to bear in mind both the inchoate state of 
medical knowledge in his time, as compared not only with today's but 
even with Galen's, and also how different was his purpose from that of 
the modern scientist. Aristotle too can surprise us with his unwavering 
adherence to teleological explanation. The idea that our makers 
lengthened and coiled the intestines in order that retention of the food 
* 69 d ????????? ?? ????? ??? ?? ???? ? ? ??????. The creation of the body and the mortal 
parts of the soul have already been referred to (pp. 309 f. above). Often quoted as the best 
example of the tension between design and necessity is the thinness of the head and its covering 
of skin, necessary to it as a seat of intelligence but rendering it more vulnerable. The gods had to 
choose between making a dull but long-lived race or one shorter-lived but better and more 
intelligent G5b-c; Lloyd, EGSc 73). 
313 
Timaeus and Critias 
might check what they foresaw as our propensity to unreasonable greed 
G3 a) has been thought merely playful, but the connexion between a 
straight gut and an insatiable appetite is noted by Aristotle in a serious 
zoological work and moreover in a teleological context.1 Many items in 
Plato's account of the body and its ailments can be traced either to 
Empedocles or to a writer of the Sicilian and Italian or Coan medical 
schools: Alcmaeon, Philistion, Diodes or a Hippocratic treatise.2 
Physiology based on physics. At the same time Plato's physiology is 
firmly based on his physics. The strength of youth and wasting of old 
age are due to the condition of the elementary triangles, fresh and 
strong in youth to cut up and absorb those that come in from outside, 
but finally loosened and enfeebled by the struggle and themselves 
broken apart by the attacks of the environment. When these reach the 
life-giving marrow, death ensues. Apart from disease or violence, life 
has its natural term, determined by the length of time for which the 
triangles can hold together.3 
Sensible qualities: body and soul.* First of all Fic-68d), Plato explains 
the sensible qualities of things in terms of the shapes and sizes of their 
1 GA 717323, repeated at PA 675318-21, b25f. (Taylor supposed that A. W3S joking too. 
See his Comm. 517.) 'An abnormally short gut is, in fact, a sufficient cause for a ravenous appetite' 
(Ogle on 675 a in Oxford trans, vol. v). The gods have saved us from behaving like Callicles's 
ideal, the ????????? (jGorg. 494 b). The prophetic function of the liver and its adjunct the spleen 
is a doubtful case G1 a~72d), but at least P. makes it an opportunity to condemn as useless the 
regular practice of taking omens from the livers of sacrificial animals G2 b). His belief in divination 
is sufficiently attested in Phdr., and the necessity for a 'sane* interpreter of the inspired ?????$ is 
typical G2a-b; for prophets who 'say many true things but don't know what they are saying' 
see Apol. 22b-c, Meno 99 c). 
2 For Empedocles see Taylor 18, and for information on both him and the medical writers 
Taylor's and Cornford's notes on the relevant passages. Cf. Taylor 410, and his commendation 
of the 'amazing erudition' of Martin's notes on this section. Note also Hoffmann's appendix to 
Zeller 11.1, §v, pp. 1070-86, 'P. und die Medizin'. Reff. to more recent discussions of P.'s relation 
to contemporary medicine are in Lloyd, JHS 1968, 84 n. 32. 
3 81 b-d and 89 b. Triangles are also referred to at 73 b and 82 d. It will be remembered that the 
deathlessness of the cosmos was ensured by making it of the whole sum of body so that it could 
not be attacked from outside (p. 279 above). The span of human life is said to be ??????????, 
which I take to mean resulting from necessity, not the planning of our makers. (In P. the  
preceding ?? ??????? need not rule this out.) For Democritus death was 'the outflow of soul-atoms 
from the body owing to pressure from the environment', and there was no immortal, part to 
escape it. (Arist. Parva Nat. 472314; vol. 11, 434.) 
4 Sight and the visible have already been dealt with at 45b~46c, but colour is treated here 
F7c-68d). On a difficulty in the account (the uselessness of experimentation), see Lloyd, JHS 
1968, 83. 
3*4 
Necessity and design in the natures of men 
constituent bodies. This is not intended as a complete explanation of 
sensation; he expressly says that the fact of sensation must be taken for 
granted, even though the creation of the body and the mortal parts of 
the soul have not yet been described.1 Being thus limited to the physical 
qualities of sensa (their ????????, 61C5) his account sounds Demo- 
critean in its materialism, though based of course on atoms of regular 
geometrical shape. The sensation of heat from fire, for instance, being a 
piercing one, is effected by small, sharp-angled, mobile particles,  
hardness by cubes with their stable bases. But whereas the materialism of 
Democritus was consistent, the sensitive soul itself being a composition 
of exceptionally small, smooth, round atoms and sensation a direct 
result of their disturbance, Plato with his incorporeal soul D6 d) has 
not that resource, and yet speaks as if the transition from physical to 
psychological motions presented no difficulty. For instance, speaking 
of sight at 45 c-d he has said that the body formed of light and the visual 
ray conveys motions 'to the whole body until it reaches the soul'. 
Similarly at 64b 'when anything naturally mobile is even slightly 
disturbed, it spreads the disturbance around, other particles (?????) 
passing it on to yet others, until it reaches consciousness and announces 
(?????????)) the effect of the agent'. If asked to attach a noun to 'the 
conscious' (?? ????????), what would Plato have said? Part (??????)? 
Faculty (???????)? As it is, the passage might have been written by 
Democritus about soul-atoms.2 The insensitiveness of bone and hair is 
ascribed solely to their material composition: they consist largely of 
earth with its motion-resisting cubes which do not pass on a shock 
F4c). Again, at 86e-8ya, on the somatic causes of psychological  
disturbances, we are simply told that certain varieties of phlegm and bilious 
humours penetrate to 'the three seats of the soul' (so the effect is not 
confined to its mortal parts) and ' mingle their vapours with the motion 
of the soul'. Plato, one may fairly say, was not at his best when trying 
to graft an atomic physical system, hitherto associated with chance or 
necessity as sole causes, on to his own very different metaphysical and 
theological stock. One's questions, which would have seemed impudent 
1 6ic6-d5· Cornford simply comments: 'The mortal part of the soul and the main bodily 
organs are reserved for the third part of the discourse, from 69A onwards* {PC 259 n.). But no 
explanation is given there of the relationship between body and soul that makes ??????? possible. 
2 For D.'s theories of sensation and the soul, see vol. 11, 430ff., 438 fT. 
315 
Timaeus and Critias 
intruders on the conversation in the Phaedo, become more pertinent 
when he sets himself up as a rival to the atomists. 
Aristotle was the first to offer, through his doctrine of form and 
matter, a non-material explanation of sensation: sensation he defines as 
reception of the sensible forms of things without their matter. Faculty 
(dynamis) and organ of sense together form one concrete individual, a 
material object endowed with the power of sense-perception. A seeing 
eye is one thing, but like every separately existing thing can be  
philosophically analysed into two components. Its material constituents and 
qualities are defined differently from its form, the faculty of sight: 
numerically one, they are different in essence {De an. 424325).  
Otherwise, he adds, what perceives would be body. This relationship exhibits 
in a single organ the relation of soul as a whole (of which sight is an 
activity) to body as a whole. ('If the eye were an animal, sight would be 
its soul', De an. 412b 18.) Whatever we may think of this, we must 
admit that we are scarcely nearer than Aristotle to understanding the 
interaction (if there is interaction) between body and mind,1 and he did 
at least see the problem and attempt to put an answer into words. Plato 
seems unaware of it. He himself, to use the modern jargon, was no 
reductive materialist.2 When we say we see with our eyes and hear with 
our ears, we mean that our psyche sees and hears through these organs 
as instruments (Tht. i84b-d). In the Phaedo it is through the body that 
the psyche becomes acquainted with the sensible world, but the wise 
psyche will resort to it as little as possible, for the body and sensation 
are a hindrance to its philosophical quest for knowledge of the Forms 
F4a-66a). There however the soul is still a unity, the intellect alone, 
and immortal. Sense-perception, emotions and desires are assigned to 
1 Paul Edwards in Mod. Intro J. to Phil? ?????. runs through the main conflicting theories 
(interactionism, reductive materialism, epiphenomenalism). Cf. W. A. Sinclair, ib. 577: changes in 
the retina cause changes in the nerves, which cause changes in the brain, 'after which, in some 
way we do not understand, we have the experience called seeing' (my italics). See now Popper and 
Eccles, The Self and its Brain A977). To Aristotle goes the credit of being the first to separate the 
physical events (which he also describes) from the experience. In De sensu D38310) he demands 
of Democritus, who had thought it sufficient to call vision a reflection of the object in the eye, 
why every reflecting surface is not capable of sight. For Plato plants, though created solely for 
the sustenance of man, have sensation and appetite, and feel pleasure and pain G7 b), but for 
Aristotle they do not, precisely because they are only acted on by the matter, not the form, of an 
external object. 
2 He strongly disagrees with the ' Giants' of Soph., who maintain that the soul itself has a body 
B47b). 
316 
Necessity and design in the natures of men 
the body. Now, in a development of its tripartite nature in the Republic? 
it has acquired mortal parts created in conjunction with its incarnation 
and occupying different parts of the body. Plato clearly wished to 
maintain the Phaedo's conception of it as belonging wholly to the 
realm of the invisible and unchanging, akin to the Forms, yet he  
sometimes comes perilously near to justifying Aristotle's criticism that he 
treats it as a physical magnitude.2 
Pleasure and pain. These are treated as sensations, and with this  
limitation are accounted for in general much as in the Philebus, but linked 
with the particulate theory of matter. Any disturbance of our natural 
state is painful, and the return to normality pleasant, if they occur 
suddenly: what happens gently and gradually is not perceived,3 and the 
nature of the change depends on the size and consequent mobility of 
the particles. One may enjoy replenishment, without having been 
conscious of a previous lack, as with the 'true' pleasures of the Philebus, 
which include those of scent E1b; cf. Tim. 6ja6). 'Bodies [sc. organs] 
formed of larger particles yield reluctantly to the agent, and passing on 
the motions to the whole, have [sic: ?????] pleasures and pains, pains 
when they are being shifted out of their normal state, pleasures when 
they are being reinstated' F4 c). These particles are presumably different 
from those just mentioned whose stability keeps their motion below the 
threshold of sensation altogether. There it was the most mobile particles, 
those of fire and air, which caused the keenest sensations. Again, the 
reason why the material sight-ray, though highly sensitive, gives us no 
pain when it is cut or burned F4d-e), is that the smallness of its particles 
enables this to happen without violence. 
1 The analogy between the partition of the soul and the organization of a city is recalled at 
7oa-b. 
2 De an. 40732-3. Cf. 406b26 and Cherniss, ACPA 393-5. It may be said that the limitation 
of the discussion to material sensa at this point is deliberate F1 c), but if so one would expect the 
problem of their relation to the perceiving soul, and the whole phenomenon of ???????, to have 
been recognized elsewhere. Note also the residence of the soul in the marrow at 73b-d, 7335. 
When the triangles of the marrow break down, the soul (presumably here the immortal part only) 
is free to depart (81 d). On the other hand, what promotes 'the motions of the souF is musike and 
philosophy (88 c). 
3 This restriction does not appear in Phil.y and is perhaps an innovation due to the particulate 
theory. 
317 
Timaeus and Critias 
Diseases of body and soul. The section on diseases is remarkable for its 
views on the physical origin of psychological disorders, disorders of the 
psyche.1 Sexual incontinence results from superfluity of semen in the 
marrow, and a man troubled with this should not be reproached as 
wicked but treated as diseased. At this point (86d7-ei) the Socratic 
maxim 'No one is willingly bad' is repeated: 'badness' is due to bodily 
defects and ignorant rearing.2 Ill health can cause bad temper,  
depression, rashness, cowardice, forgetfulness and stupidity. If men with these 
initial disadvantages live in a badly governed community, the chances 
of amendment are still further reduced. This does not, however, imply 
a rigid determinism. Everyone's condition may be improved by  
education and training, study, and choice of pursuits, and these are the joint 
responsibility of parents, community and the individual himself. (But 
these, adds Plato at 87b 8, are subjects for another occasion.) His aim, 
as always, is balance, symmetry, due measure and proportion. Soul 
(mind) must not be too strong for body nor body for soul. The remedy 
(which is relevant to the present discourse) is to give plenty of exercise 
to both. The naturally studious must not neglect physical training, nor 
the athlete mental culture. Even in sickness, exercise and diet are, 
except in the gravest cases, better than the use of drugs, which only 
upset the natural course of the distemper. 
As for the soul itself, this calls, he says, for only a glancing reference 
to what he has said before. The three forms of soul in us, with their 
several motions, must be kept in proportion. Over-exercise of the 
appetitive and ambitious parts will fix our thoughts on mortal things, 
and the divine spark will fade. Our duty is, through the pursuit of 
knowledge and true wisdom, to cherish the divine and immortal part, 
placed at the summit of the body and raising us towards our heavenly 
kindred, a guiding spirit given us by God. In this way we shall attain 
the fullest immortality possible for the human race. Everything  
flourishes by exercising the motions proper to it, and the proper motions of 
1 On the extent of P.'s originality here see Lloyd in JHS 1968, 87. Note also the highly 
somatic description of emotions at 7oa-d. 
* ????? probably refers primarily, as Taylor suggests, to bodily regimen, though it can cover 
education too. Cf. ????? ????*????$ at 44 c. For the Socratic paradox see vol. in, 459-62, and 
for P.'s retaining it while still maintaining a distinction between ignorance (cured by education) 
and wickedness (cured by punishment) see p. 126 n. 3 above. 
318 
Necessity and design in the natures of men 
the divine part of us are the revolutions and thoughts of the cosmos. 
By studying these we may repair the damage done to our own circuits 
at birth and bring our minds to resemble the Mind which is their object, 
so achieving the best life offered to us by the gods, now and for ever. 
APPENDIX 
The narrative order 
By his own confession C4c), Timaeus's narrative has much in it of the 
casual and random, and follows neither the order of events nor a logical 
order. Beginning with the reasons for the world's creation and the nature of 
its model, he passes to consider the making first of its body, then of its soul, 
with the warning that as the soul is of higher worth than the body, so also it 
preceded it in the order of generation C4b-c). There follows the creation of 
the heavenly bodies (making time as we know it possible), of the stellar 
gods and mankind, including the nature and destiny of human souls and the 
effect on them of incarnation, and ending with an account of the purpose and 
mechanism of sight. 
So far, he says D7 e), he has spoken mainly about the work of Reason. 
Now he must begin all over again D8 d-49 a) and tell of the subordinate 
cause, necessity, and its effects 'before the heavens came into being' E2d). 
(These have therefore been taken before the creation in the present chapter.) 
There follows the description of the pre-cosmic chaos, the Receptacle and 
its tumultuous contents, 'without proportion or measure', which brings him 
back in a circle to the work of Reason, which was precisely to introduce the 
missing measure and proportion and produce the four bodily elements in 
their present distinct form by organizing them into microscopic particles of 
regular geometrical shape. So at 53 c he picks up again the theme of the body 
of the cosmos, which he left at 34 b. He describes the structure, varieties and 
combinations of the elements, and the sensible qualities that result from them. 
(We are asked at 61 d to assume the faculty of sensation and its fleshly organs, 
though they have not yet been described, because one cannot, unfortunately, 
talk of both topics at once.) 
The natures and effects of the physical elements are still partly determined 
by necessity. Timaeus now 'makes another fresh start' F9a) to remind his 
audience that' these things' were in confusion until God bestowed measure 
and proportion on each one and on their mutual relations, so far as their 
nature allowed.1 The rest of the dialogue (from 69 c) tells how necessity was 
1 Cornford and Lee divide the dialogue into three main parts, classifying ?,??-^?? as the work 
of Reason, then the whole of 476-693 as the work of necessity, and only what comes after that 
319 
Timaeus and Critias 
subordinated to intelligence in the creation of human bodies and the mortal 
parts of soul, which God delegated to the minor gods, his offspring. It deals 
in clinical detail, from Plato's teleological point of view, with bodily parts 
and organs and their functions, and the cause and prevention of diseases. 
The dialogue ends with a hasty tailpiece on the origin of women, the 
physiology of sexual reproduction, the lower animals, and a final sentence 
summing up the excellence of the universe as a visible, perceptible god. 
as the collaboration between Reason and necessity. They take * All these things were so constituted 
of necessity* at 68e to refer to what has been described in the previous section, which however 
relates to the cosmos after the Demiurge had ordered the elements into regular geometrical 
shapes. Both scholars ignore the temporal force of ???? ... ????? F8 a i and 3). All these things 
owed their condition to necessity until the Demiurge took them over and used their causes as 
accessories in creating the cosmos. In what preceded we were shown the limitations still imposed 
by necessity on the elements as constructed by Reason. 
32O 
? 
THE LAWS1 
Introduction 
Authenticity and date. 'Jowett opens with a proof, unnecessary today, 
of the authenticity of the Laws . . . The authenticity of the work ... is 
no longer questioned.' So Jowett's editors in 1953, and most would 
agree today. The Laws contains infelicities of style, irregularities of 
syntax verging occasionally on incomprehensibility, repetitions and 
internal inconsistencies which some nineteenth-century scholars, on the 
familiar' unworthy of Plato' argument, attributed to heavy posthumous 
editing on the part of Plato's pupil Philip of Opus.2 External evidence 
they found in (i) D.L. 3.37: 'Some say that Philip of Opus transcribed 
(?????????) the Laws, which were on wax.3 They add that the Epinomis 
is his work'; (ii) Suda, s.v. Plato: '[Philip] divided the Laws of Plato 
into twelve books, and the Epinomis is said to be his work.' These 
statements, even if based on a sound tradition, do little to substantiate 
a charge of extensive meddling with the Laws, which they clearly 
distinguish from the Epinomis, and which was known to Aristotle 
himself as Plato's.4 Nowadays it is recognized that faults of the type 
found are more naturally explained by the unrevised state in which 
1 For bibliography see T. J. Saunders, Bibliography on P.*s Laws A976). 
2 Taran has performed the valuable service of collecting and evaluating the testimonia  
concerning Philip's work and activities. See his Academica A975), § in, pp. 115-39. M. Krieg {Uberarbei- 
tung d. 'Geset^e' 1896) was one who argued that Philip made serious alterations and additions to 
Plato's text. 
3 Wilamowitz {PI. 1, 655 n. 1), reasonably surmising that no one would write a work as long 
as the Laws on wooden tablets, took these words as metaphorical, meaning 'im Konzept' ('from 
the rough draft', Lesky, HGL 538). He was following Bergk, who explained the figure as taken 
from a model coated with wax for casting in bronze by the cire-perdu process. (See Morrow, PCC 
515.) Taran however {Academica 130 n. 542) doubts whether the expression is metaphorical. The 
most judicious estimate of the part played by Philip is that of von Fritz in RE xxxviii. Halbb., 
2360-66. Raeder {PPE 398 f.) went so far as to doubt the actual fact of a posthumous editing. 
4 See Pol. 1265 b 5 and Taran, Academica 131 n. 548. As evidence of early circulation many 
scholars confidently cite Isoc. Phil. 12, which however speaks only of * laws and constitutions 
written by the Sophists'. Isocrates may sometimes have hinted at Plato under the guise of a 
Sophist (vol. iv, 3iof.), but the reference here could as easily be to Protagoras, who according 
to D.L. (9.55) wrote both a Laws and a Constitution, or Antisthenes (D.L. 6.16). 
22 
321 
GHG 
Laws 
Plato had to leave this work of his old age, and their retention as a mark 
of the scrupulousness with which Philip carried out his work of  
copying. On the critical, or subjective, side, Ast A778-1841) wrote: 'One 
who knows the true Plato needs only to read a single page of the Laws 
in order to convince himself that it is a fraudulent Plato that he has 
before him here.' Brochard in 1926, after claiming that Gomperz in 
1902 had definitely established its genuineness, goes on to say that in spite 
of some lenteur and negligence, the unity of plan, the vigour of the 
general conception, the beauty of certain pages and the perfection of the 
ensemble show us Plato still in full possession of his genius.1 For myself, 
coming to it (as the reader may grant) after a fairly close study of all the 
other dialogues, I feel no doubt that, to adapt the ancient critic's verdict 
on the Odyssey, it is a work of old age, but definitely—even if (as the  
content may occasionally make one think) regrettably—the old age of Plato. 
The only early external evidence for the date of the Laws is Aristotle's 
statement in the Politics A264 b 26) that it is a later work than the 
Republic. Internally, the defeat of the Locrians by the Syracusans  
mentioned at 638b is thought to be that of Dionysius II in 356 B.C.,2 which 
means that bk 1 at least was written when Plato was over seventy.Evidence 
that bk 12 too was written after this date was seen by Grote at 944 a. 
(See his vol. 111, 443 n. q.) Plutarch (Is. and Os. 37of.) speaks of Plato 
being 'already an old man' when he wrote the Laws, and there is much 
in the tone of the work to suggest that he wrote it after the failure of his 
last visit to Sicily in 360.3 If so, the composition of so long and detailed 
a work may well have been still occupying him at the time of his death, 
as tradition says.4 Yet his disappointment never destroyed his faith in 
1 Ast quoted by Harward, Epin. 34; Brochard, Etudes 154, referring to Th. Gomperz's 
decisive article 'Die Composition der Gesetze' in S.-B. Wien, ph.-hist. Kl. 145 A902). Not to 
leave modern technology unmentioned, according to Morton and Winspear (Grk to the C. 13, 
78 f.) bks 5 and 6 are stylometrically un-Platonic, revised by another, possibly Speusippus. 
2 Or 352: see Taran, Academica 132 n. 554. 
3 This is largely a matter of general impression, but Ep. 3, 316a may indicate that his work 
with Dionysius II on that visit provided the prototype for the 'preambles' of the Laws. (So 
Harward, Epp. 179f. and others. For these ???????? see p. 336 below.) 
4 Morrow, in his essay 'Aristotle's Comments on P.'s Laws' (During and Owen, A. and P. 
in Mid.-4th Cent. 145-62), argued that when Aristotle wrote Pol. bk 2 he only knew bks 3-7 of 
the Laws, and that this must have been in P.'s lifetime, before he wrote the rest. Ryle in P.*s P. 
(89, 257-9) makes much of this thesis of a 'Proto-Laws\ but see T. J. Saunders in Rev. Beige 
de Philol. et d*Nist. 1967, 497. Though Friedlander agrees that parts of the dialogue could be 
older than others, the whole of his ch. xxxi reads as a \ indication of its essential unity. 
322 
Introduction 
what could be achieved if a young man, talented, brave and with the 
gift of self-control, should gain dictatorial powers and be fortunate 
enough to have at his elbow a legislator of the right sort. Teachable 
tyrant and wise adviser still constitute his recipe for a happy and well- 
governed state G09e-iob). As we know (see vol. iv, 23f.), several 
members of the Academy were invited by existing states to draw up or 
reform their laws, and in the Laws Plato lays down the principles on 
which they should act and offers a model of an actual constitution and 
legal code for their guidance. 
Characters and setting. Three old men, as they frequently call  
themselves though still capable of walking many hours in the heat of a 
Cretan midsummer's day F83 c), set out from Cnossus for the cave and 
shrine of Zeus on Mount Ida.1 They are Clinias, a native Cretan, 
Megillus a Spartan, and an anonymous Athenian2 who leads the talk. 
The little that is told of the other two is designed to indicate that they 
will make sympathetic and attentive listeners. Clinias has family  
connexions with Epimenides, the Cretan seer who encouraged the Athenians 
with his prophecies about the Persians, and has himself inherited  
feelings of goodwill towards Athens. As for Megillus, his family at Sparta 
were proxenofi of Athens which he had been brought up to regard as a 
1 This is the cave to which the legendary Minos is said to have ascended periodically, bringing 
back laws which like his predecessor Rhadamanthys he claimed had been delivered to him by 
Zeus—falsely according to Plato's contemporary Ephorus. See Strabo 10.8, p. 476. For the 
Cretan therefore it would be a pilgrimage, and it makes the walk an especially fitting occasion for 
a discussion on lawmaking. On the Idaean Cave as their destination rather than other caves 
sacred to Zeus, see Morrow (p. 324 n. 3), 27 f. 
2 At Pol. 1265an Aristotle makes a general criticism of 'all the discourses of Socrates', in 
which he includes the Laws. (He does not mention the name Socrates three times as the Oxford 
translation does!) The criticism seems intended for all Plato's dialogues, and the name may have 
slipped in inadvertently as Cherniss suggested. (See Morrow in During and Owen 146 n. 3.) 
At 1266b5 it is replaced by 'Plato in the Laws*. Some have thought however that A. is referring 
to an earlier version in which Socrates himself was the speaker. So Ryle, P.'s P. 258. Morrow 
however, whose suggestion it was that the Laws was only partially extant when A. wrote Pol. 2 
(PCC inn. 44), does not believe this. See During and Owen 146. I cannot believe that P. ever 
made Socrates the speaker. Apart from the un-Socratic (sometimes even anti-Socratic) character 
of much of the work, a Socrates tramping the Cretan mountains is unimaginable (see especially 
Crito 52b and Phdr. 230c-d), and no one, I think, has suggested, on no evidence at all, that the 
scene too was altered during composition. 'The shift of scene to Crete and the introduction of the 
Athenian are to be thought of as a single creative act' (Friedlander, PI. 1, 362 n. 9; cf. in, 388 f.). 
3 Citizens of one state who represented the interests of another in their own. The connexion 
was usually hereditary. 
323 
22-2 
Laws 
second fatherland F42 b-d). He is not especially intelligent and  
maintains the Spartan reputation for brevity, to which he refers at 721c 
Clinias has more to contribute, but both are completely overshadowed 
by the Athenian, Plato's own mouthpiece.1 He is allowed long stretches 
of monologue, and there is no real argument. Together, the participants 
symbolize the system to be expounded, largely Athenian in origin, 
stiffened with Dorian, and especially Spartan, discipline.2 The setting 
is not entirely forgotten as the dialogue proceeds (cf. 683 c, 722 c), but 
lacks that integration with the talk which the Phaedrus so skilfully and 
delightfully displayed. 
Plan of chapter. The twelve books of the Laws include several devoted 
entirely to the exposition of a constitution and a legal code dealing with 
everything from subversion and treason to neighbourly disputes over 
drainage, enticement of bees or filching of fruit. To follow these in 
detail is neither possible nor, fortunately, necessary since their thorough 
analysis by Morrow.3 But before the model city and its laws are reached 
there are three books of an introductory conversation on such topics 
as legislation in the Dorian states, the lessons of history, types of 
constitution, education and the arts. The recommendations for the city 
itself, apart from externals like town-planning G78a~79d, 848 c-e), fall 
into two halves, the establishment of offices with the methods of  
election or appointment, and promulgation of the laws which the elected 
officials will be expected to enforce G35 a). The necessarily curtailed 
treatment here will be divided into four main sections: A) The  
introduction (bks 1-3), B) The city of the Laws, C) Life in Plato's city, 
D) A few more general philosophical points. One aim will be to keep 
the other dialogues in mind and see how much of his earlier views Plato 
1 Gigon's observation, that it is a puzzle to know what an Athenian and a Spartan are doing 
in Crete (Mus. Helv. 1954, 207), is just, whether or not one accepts all the conclusions he draws 
from it. 
2 Some bibliography on P.'s attitude to Sparta will be found in Tigerstedt, Legend of Sparta, 1, 
544f., n. 202. His own estimate of Sparta's influence on the Laws differs markedly from Morrow's. 
3 G. R. Morrow, P.*s Cretan City (i960; hereafter simply 'Morrow'). See also Gernet, 'Les 
Lois et le droit positif, Bude ed. pp. xciv-ccvi, H. Cairns, P. as Jurist, ch. xxxi of vol. 1 of 
Friedlander's Plato, and for earlier refF. Leisegang in RE 2514. Saunders's translation has a full 
analytical table of contents (pp. 5-15). Special mention must be made of Jerome Hall's essay on 
'Plato's Legal Philosophy' (Indiana Law J. 1956), a valuable appreciation of P.'s whole theory 
of law from the standpoint of a modern legal and political theorist. 
324 
Introduction 
has abandoned or retained. I hope that in spite of the necessary  
compression and omissions, what follows will succeed in preserving a 
proper balance. 
(i) Introductory conversation {bks 2-3) 
Aims and methods of education, with special reference to the use of drink 
{bks 2 and 2). The laws of Crete and Sparta (say Clinias and Megillus),1 
with their emphasis on courage and physical endurance, were devised 
on the assumption that cities are in a continual state of war. Conflict in 
fact is the keynote of life, from cities down through villages to families 
and individuals. ' And within an individual?' asks Plato (for surely one 
may give the Athenian this name), introducing at once the familiar 
theme of internal tension between a man's better and his worse self, of 
mastering or being mastered by oneself {Gorg. 491 d, Rep. 430e, Phdr. 
237 d-e), in short of the virtue of sophrosyne. This internal struggle 
occurs also in states, where the worse element may rise to the top :2 but 
laws should aim at reconciliation and peace, not war, and encourage all 
virtue, not just a part of it (courage), and that the least.3 In any case, 
courage is shown not only in facing dangers and enduring pain but also 
in withstanding desires and pleasures—a reminder especially of Laches 
191 d-e, but also Rep. 413 d-e. Spartan training hardened its youth in 
dangers and pains but offered no opportunity of testing their resistance 
to the seductions of pleasure. Quite right too, thinks Megillus. He has 
been shocked by the sight of drunkenness at parties and festivals in other 
states. Drink simply weakens a man's resistance to temptation, and its 
prohibition in Sparta has been a source of strength. The Athenian 
replies with an extraordinary eulogy, in several parts and at tedious 
length, of properly conducted drinking parties, not stopping short of 
intoxication, as a beneficent educational influence and test of moral 
stamina.4 The argument is this. Wine increases passions and weakens 
1 For the traditional connexion between Cretan and Spartan Laws see Hdt. 1.65.4. 
2 The parallel is drawn all over again at 689 b. 
3 Cf. the remarkable passage in bk 6, 803 d-e, on the avoidance of war (cited on p. 352. 
below). 
4 The main passages are 639c~4id, 64$d-$ob, 67ia-72d, 673d-74C For P.'s sake one must 
hope those are right who attribute its length to the lack of opportunity for revision and excision. 
There are also a few other points to notice which may modify the impression given here. The 
need for ????? ?????????? F71c) reflects the practice in the Academy. (See vol. iv, 20 f.) In the 
325 
Laws 
judgement. It can make men childish and incapable of self-mastery. So 
something to be avoided at all costs? Not if it means accepting  
temporary incapacity to gain lasting good, as with some medical treatment 
and exhausting exercise. Fear is of two kinds: (i) of pains and the like, 
B) of infamy, also called a sense of shame. This must be cultivated, and 
its opposite, shamelessness, avoided as a great evil. A man learns to 
overcome fear by being brought (as in the Dorian states) into controlled 
contact with danger and pain. Similarly he must learn to overcome the 
temptations of desire and pleasure by experiencing them under  
controlled conditions. The master of the revels must be of mature age (over 
60) and remain sober F40d, 671 d-e). Wine is a safe test of character. 
To discover whether a man is a cheat or a sex maniac, one need not have 
business dealings with him or put one's wife and children at risk: in his 
cups he will reveal his true nature. The educational object is to produce 
men like Socrates, who can expose themselves to the risks of drink and 
other temptations without losing self-command. 
The importance of pleasure and pain in human life can scarcely be 
exaggerated. They may be ' witless counsellors', but are also the cords 
by which the puppet Man is manipulated. The study of law is almost 
entirely an investigation of pleasures and pains, and law itself may be 
defined as ' the public decision of a city on the relative merits of pleasure 
and pain V No one would voluntarily act in a way that brings him more 
pain than pleasure F63 b), but fortunately the virtuous life is the most 
pleasant. So the Athenian argues in bk 2, and in bk 7 G326-333) that it 
is natural to human beings to feel pleasures, pains and desires, and the 
best life is praiseworthy just because, if people will only give it a trial, 
it ensures a predominance of pleasure over pains for the whole of life.2 
Platonic society no one under 18 may touch wine at all, and no one under 30 get drunk F66 a). 
The gift of Dionysus was not intended to madden us but to produce a sense of shame or reverence 
(?????) in the soul and health in the body F72 d). Its use must be state-regulated for these 
purposes, not solely for amusement. It is forbidden altogether to slaves, serving soldiers,  
magistrates during their term of office, jurymen and ships' steersmen; and 'licensed hours' are restricted: 
no one may drink until evening. In view of all these regulations viticulture will be strictly limited 
F73e-740. 
1 See bk 1, 6440-453, 636 d. (Law will have other definitions before the Laws is finished.) 
This is one indication that most of the Laws is concerned with 'popular virtue', not the  
philosophic virtue to which the calculation of pleasures and pains is irrelevant. They simply 'nail the 
soul to the body', escape from which should be its highest endeavour (Pho. 69a-b, 83d). 
2 In bk 7, 792 c-d, the Athenian declares that the right life must not pursue pleasures nor shun 
pains altogether, but accept a neutral state, which indeed is v, hat the gods enjoy. This apparent 
326 
Introductory conversation 
These considerations lead, at the beginning of bk 2, to a definition 
of education in general as inculcation of the right attitude to pleasures 
and pains, loving the good and hating the bad. It is best inculcated 
through music and dance, by which the natural restlessness of all young 
things is translated into the distinctively human gifts of rhythm and 
melody. The moral effect of these, for good or ill, is so strong that, 
as in the Republic, the composition of tunes and songs must be strictly 
controlled by law, and innovations frowned on.1 Apollo's and the 
Muses' gifts are for our recreation and delight, but it is through play, 
song and dance that children can be educated to accept what the law 
approves. The lesson of the Republic remains valid.2 Clinias claims that 
this is recognized only in Crete and Sparta, whose example other Greek 
states would do well to follow, and the Athenian innocently assumes 
his agreement that a man with every external advantage including 
tyrannical power is wretched and unhappy unless he is also just and 
good. Clinias does not agree. Such a life is morally reprehensible 
(aischrori) but not bad, unhappy or unprofitable (kakoii). So we are back 
at the argument with Polus in the Gorgias and the Socratic point, which 
the Athenian goes on to defend, that the most righteous life is also the 
pleasantest.3 This is the truth, but even if it were not, a lawgiver's only 
consideration should be what beliefs in the young will most benefit the 
state. Again the ambivalent attitude to truth which we found in the 
Republic: the philosopher must pursue truth exclusively, yet the 
4 medicinal' use of spoken falsehood is recommended to the founding 
fathers (vol. iv, 456-c>).4 
inconsistency can hardly be understood apart from the analysis of pleasure and pain in PA/7., 
especially 32c See p. 224 above. 
1 Later, in bk 3, 7003-701 a, P. sees the first signs of the decline of Athenian morale in the 
innovations and decadence of Athenian music. The strict censorship of poetry, drama and music 
recurs in bk 7 at 801 c-d. See also 817a and bk 2, 660a. Friedlander (PL in, 560 n. 29) has noted 
the close verbal parallel between 801 e and Rep. 607 a. To forbid innovation was a Spartan  
tradition according to Plut. Inst. Lac. 238c. 
3 See vol. iv, 450??. for this and the general Greek association of aesthetic with moral values. 
On song and dance in the Laws, and their general importance in Greek life, see Morrow 302-18. 
3 For the Gorg. see vol. iv, 288f., and cf. Cr'ito 49b. (Wrong-doing is ????? as well as ???????.) 
And of course the happiness of the just is a leading theme of the Republic. 
4 For truth as, in the Laws, the highest good for men and gods alike see 730 c. The late C. D. 
Broad, one of the gentlest of men, expressed his full agreement with Plato that the political use 
of myth or fiction by governments was justified as an instrument for the promotion of good 
conduct. See The Mind and its Place in Nature 511 f., of which I was reminded by Bambrough 
in RTG^. 
327 
Laws 
To return to music, choruses for song and dance will be divided 
according to age: children, under thirties, and the middle-aged from 
thirty to sixty. Since the last-named would feel a natural embarrassment 
at being detected in such activities,1 they may perform at private 
parties, softened up and rejuvenated by their patron god Dionysus, 
tipsy and merry F71 a-b, 672 a) under the eye of a supervisor over 
sixty. A regrettable casualty of Plato's old age seems to have been his 
sense of the ridiculous.2 The idea of this ' Dionysiac chorus' shocks 
Clinias, but the Athenian is undeterred. Its members will be better 
educated than others, including the song-writers themselves, to  
understand what rhythms, tunes and words are not only enjoyable but correct 
and salutary. Music and dance are mimetic arts F68 b), and besides 
pleasure and charm (?????) should aim at truth and utility. To judge a 
piece the elders must penetrate to its essence (????? 668 c) and aim. So 
equipped, they will both enjoy a harmless pleasure themselves and 
instil virtuous habits in the young. Practised in this way, singing and 
dancing are not only a delightful recreation but may be equated, so 
Plato now claims, with the whole of education, divided according to 
Greek custom into ' music' and ' gymnastic \ In view of the importance 
assigned to all forms of gymnastic in the Dorian states (see 625 c), its 
virtual restriction to dancing here is surprising, and of course Greek 
musike meant not only music but a whole literary education.3 
Bk 2 ends with a return to the subject of drink, its uses and  
regulation, and bk 3 begins:' So much for that. What are we to say about the 
origin of political organization?' There is no apparent connexion with 
the first two books. In the form in which we have the Laws, it looks as 
if Plato was dissatisfied with this treatment of education as a product of 
song, dance and intoxication. If so, one may sympathize. England  
regarded the disquisition on drinking as a general introduction to  
education^ and such indeed Plato calls it; but as it stands it is an odd, unsatis- 
1 There is some doubt about their age. At 670b they are 'the fifty-year-olds'. 
2 Aristotle, whose Politics contains many echoes of the Laws, assigns dancing to the young, 
and allows those of maturer age to sit as judges of it (i34ob35-t)). Nevertheless, apart from the 
festive details, the three choruses of boys, men in their prime, and the old were no fancy but 
existed in Sparta. Plutarch quotes from their songs (Rep. Lac. 238b, Lye. 21). 
3 Cf. Rep. 376c Admittedly music did play a larger part than with us, owing to the Greek 
belief in its moral effects. Still, we may be relieved to learn in bk 7 that education has other sides 
as well. 4 Laws vol. 1, 340, and cf. ion. 1. 
328 
Introductory conversation 
factory and inordinately long one.1 The best points made in bks ? and 2 
are all repetitions of familiar Socratic or earlier Platonic tenets. Such are 
the need for self-mastery (where symposia could have been brought in 
briefly as an illustration); the difference between knowledge and true 
belief F32 c); a hierarchy of goods as goods of soul (intellectual and 
moral),2 of body (health), and lowest of the three classes, material 
possessions F97b); courage as not only physical but moral; the 
centrality of pleasure and pain in human life; the assimilation of a man 
to what he enjoys F56b, cf. Rep. 500c); the rigorous control of poetic 
subject-matter, education through play, the happiness of the good and 
misery of the wicked, the need to know what a thing is, and the  
equation of its essence with its purpose.3 
The unity and multiplicity of virtue. One of these previously treated 
topics deserves special mention for its suggestion of a deliberate  
development from beginning to end of the Laws, and hence of its  
fundamental unity in spite of any disorder and untidiness of arrangement due 
to its unfinished state. I mean the relation of virtue as a whole to its 
parts. In the Protagoras Plato was at pains to show that the so-called 
virtues were not separate traits but only different aspects or 'parts' of 
a unity, the single virtue which was knowledge, source of all right  
conduct, whether in dangers (courage), temptations (self-mastery),  
relations with the gods (piety) or one's fellow-men (justice). To possess one 
was to possess all. In bk 1 of the Laws the Athenian speaks about 
courage and the other kinds or species (eide) of virtue and virtue as a 
whole, and of treating courage first without the rest, with no suggestion 
that their relationship presents a problem. This is obviously right in its 
context. His aim is practical, to counteract the Dorian over-emphasis on 
physical courage, and his honest companions are no clever Sophists to 
relish a philosophical discussion of the unity of virtue. He therefore 
takes the commonsense view: courage is a species of virtue, attainable 
on its own, but a legislator ought to inculcate the whole F3od, 705 d). 
1 It is only fair to add that he is aware of this and has his excuse. See 642a, and also p. 382 
n. 2 below. 
3 To understand ???? ??? ?????? F31 cj)f which offended Gigon (Mus. Helv. 1954, 225), 
I suggest one should look at Meno 88 b 7-8. 
3 ??? ???' ????? . . . ??? ??????, ?{ ???? ???????? F68 c). 
329 
Laws 
One may use ' parts' or any other word provided the meaning is clear 
F33 a). In bk 3 F96b) the possibility is envisaged that a man may be 
very brave yet wicked and licentious, exactly what Protagoras  
maintained in the Protagoras C49 d) and the Platonic Socrates denied. When 
however in the last book Plato is describing the educational  
requirements of his supreme committee, the Nocturnal Council (also called 
Guardians and corresponding to the Guardians of the Republic in having 
not simply true belief but actual knowledge), he does put the question 
and his mature philosophy comes into view (963 c-d). We speak of four 
kinds of virtue, as if each was a separate thing, yet we call them all by 
one name, 'virtue', as if they were not many but one. How is this? It is 
easy to see the differences, e.g. courage differs from wisdom by being a 
purely natural faculty found in wild animals and children. It does not 
need reason (logos), without which no soul can become wise.1 But in 
what sense are they a unity? No answer is given (this was not the place 
for it), but it is said that to answer it calls for training in the dialectical 
method of collection and division, a method with which Plato has 
wrestled in the later dialogues from Phaedrus on. This will concern us 
further when we come to consider the purpose and training of the 
Nocturnal Council (pp. 371 f. below). 
The lessons of history (bk 3). How did political communities arise? 
Imagine life starting afresh after a great flood.2 All civilization has been 
engulfed, and the only survivors are a few mountain herdsmen, good 
people but ignorant and illiterate. From separate families patriarchally 
governed they gradually form larger units, descend into the foothills 
and progress from stock-raising to agriculture. At this stage they would 
choose primitive legislators, to unify the different family traditions into 
a common code and also to appoint one or more governors to administer 
the new rules of behaviour, creating the first monarchy or aristocracy. 
(So the earliest legislators are already to be separated from the execu- 
1 This of course departs from earlier works, where P. follows Socrates in equating the virtue 
of courage with knowledge, and judgement based on knowledge. Unthinking rashness is no 
virtue, and may lead to harm. For this view see vol. in, 451-3 (Socrates), iv, 128 n. 1, 219-21, 
228 (Prof.), and Meno 88e, Rep. 430b. Here P. is speaking of the 'demotic' virtues. 
3 For P.'s uses of the belief in recurrent natural catastrophes see Tim. 22c-e, Critias io4d-e, 
Pol. 273 a and Guthrie, In the B. 65-9. 
330 
Introductory conversation 
tive.) With the third stage, the foundation of cities in the plains, Plato 
moves into the light of history, starting with the fall of Troy, the exile 
of the Achaean victors from their homes, and their return under the 
name of Dorians.1 This brings the trio back to their starting-point, the 
institutions of the Dorian states, the object being to appraise their merits 
and defects, to ask why some have survived and others not, and in 
general no less than to discover what changes would ensure the welfare 
and happiness of a city F83 b). 
From here Plato goes on to explain, on his own principles, the failure 
of the alliance between the three chief cities of the Peloponnese, the 
survival of Sparta, the defeat of Persia and the subsequent decline of 
Athens. The first was due to concentration on military strength and 
physical courage at the expense of a proper balance between the virtues. 
Undue subservience to the pleasure-pain standard and neglect of what 
is fine and good ruins state and individual alike. A soul whose grosser 
elements oppose the faculty of knowledge and judgement is like a city 
in which the mob refuses to obey its rulers and the laws. Internal  
concord is more important than professional competence, and no man who 
lacks it should be entrusted with government. In the Peloponnese (as 
usually in absolute monarchies) the rot started from above, in the 
discordant souls of the three kings, puffed with pride and greed,  
breaking their oaths and the laws. The remedy, discovered by Sparta alone, 
is constitutional reform. Irresponsible power inevitably corrupts, and 
the solution lay in Sparta's division of powers, exemplified in the dual 
kingship, the council of elders and the ephorate, introducing the 
invaluable element of measure and proportion. 
Need for a mixed constitution. There are two extreme forms of  
government, absolute monarchy (or tyranny) and democracy, represented by 
Persia and Athens respectively. All others are modifications of these. 
For a state to be free, united and wise, it is absolutely essential that it 
combine elements of both.2 By robbing the people of all liberty the 
1 For an appraisal of P.'s view of Greek history, see R. Weil, L* * Archeologie* de P.f perhaps 
with the review by Kerferd in CR 1961, 30 f. 
3 693 d. Cf. the warning against ???????? ????? at 693b, and bk 6, 756? ????? ... ?????????* 
??? ???????????? ?????????. For the theory of the mixed constitution, originated by P., see 
Morrow ch. x, von Fritz, Theory of the Mixed Constitution in Antiquity, and other works referred 
to by Morrow, 521 n. 3. 
331 
Laws 
despotic and self-indulgent kings of Persia destroyed any sense of  
community in the state. In their wars of greedy aggression, they ended by 
being unable to rely on the loyalty of their own soldiers. The history of 
Athens on the other hand shows that excess of liberty is no less  
debilitating. When they threw back the Persian invaders, the people were the 
Voluntary slaves' not of a tyrant, but of the traditional laws and their 
own moral sense (aidos).1 The falling-off from this admirable state 
showed itself first of all in music, with composers breaking the  
established rules and confusing the different genres to please uneducated and 
noisy audiences.2 Thence it extended to general disregard for law and 
an impudent refusal to listen to good advice. In Sparta and early  
fifth-century Athens we saw how in a modified form authoritarianism and  
freedom alike lead to success and prosperity, but Persia and contemporary 
Athens show the disastrous effects of either when carried to extremes. 
At this point the Athenian wonders what test there could be of the 
practical utility of their discussion, and Clinias suddenly reveals that 
he has been put on a commission to draw up laws for a new colony in 
Crete. Its population will be drawn from various cities, but Cnossus has 
been charged with the planning. He suggests that it will help him if they 
now plan an imaginary state on the basis of the points so far made. 
B) The city of the Laws 
Was it intended to be realiied in practice? The idea of starting a brand- 
new state with prefabricated constitution and laws is less familiar to 
Britons than it was to the Greeks, with their habit of founding  
politically independent colonies or offshoots of existing city-states. One 
might say that when the United States ceased to be a British colony it 
became a colony of Britain in the Greek sense. We have seen, too, in 
connexion with the Academy, how existing states would call in expert 
legislators from outside to reform their political and legal systems.3 But 
how far did Plato intend or hope that his scheme for Magnesia* might 
1 Cf. the answer of Demaratus to Xerxes, Hdt. 7.104 (vol. in, 79). P.'s brief account of the 
spirit in which Athens met the Persian peril is in his best vein as a writer. 
3 7003-701 b. 700 c closely parallels Rep. 492 b-c. 
3 Vol. iv, 23, and cf. the pan-Hellenic colony of Thurii founded by Pericles (vol. in, 264). 
In some cases (cleruchies) the colonists retained citizenship of the mother-state. 
4 In the later books the citizens are several times referred to as 'Magnesians' (first at bk 8, 
86oe). 
332 
The city of the Laws 
be realized in practice? In strong contrast to the Republic (a literal 
U-topia or Erewhon) it is located in Crete, named, and its site described: 
the deserted site of an ancient city Magnesia, some ten miles from the 
sea, with good harbours, self-sufficient in natural produce but with no 
surplus for export. The availability of timber for ships, and the  
advantages and disadvantages of a population of mixed origins, are also 
discussed.1 In bk 5 Plato claims, with obvious reference to the Republic, 
that the ideal would be the abolition of all private ownership, in wives 
and children as well as goods. Perhaps some gods or sons of gods live 
under such a rule, but we must be content with human nature as it is, 
4for we are addressing ourselves to men, not gods'.2 We are no longer 
in the age of Cronus, when men were governed not by other men but 
by gods.3 
In the Republic many things are left in the air which it would be 
essential to explain if it had been intended as practical politics. How the 
new regime would ever be brought into force is a question never 
seriously faced. In Laws 6 this is elaborately provided for. A mixed 
company of settlers from different states, strangers to each other, will 
not be immediately in a position to elect the best men to office G51 d).4 
The Cnossians must therefore appoint a commission of 200, selecting 
100 of the best and most mature among themselves and 100, as far as 
possible similarly endowed, of the settlers, to choose the first officials. 
Pre-eminent among these will be the 'guardians of the laws', a board of 
37 (initially 18 Cnossians and 19 settlers), aged between 50 and 70 
G5 5 a)> who besides their primary duties will keep the register of private 
property, each citizen having to declare his own. As time goes on and 
the constitution is established, this body, like the Council and other 
offices, will be recruited by internal election on mainly democratic lines, 
for which elaborate rules of procedure are laid down. The idea of a new 
1 In bk 3. We are here concerned only with the realism of P.'s picture. For a possible connexion 
with the Asiatic Magnesia on the Meander and an actual site at the W. end of the plain of the 
Messara in southern Crete see Morrow 30 f., which may be called a defence of Plato's local 
knowledge against Wilamowitz and Taylor. (See Taylor's PMW 464.) 
2 739c-e, 732e, 807b. At 739 d ??? ???? ?????? ????????? ??? ??????????? matches Rep. 
464 a. This remained a permanent ideal for P. Note that it is in both passages explicitly 
connected with the community of wives. 
3 7i3c-d. This feature of the Cronus myth is repeated from Pol. 271? (pp. 181 f. above). 
4 Against the idea that 7513-5 5 b represents a conflation of two originally separate sets of 
proposals (Wilamowitz, revived by Morrow) see Saunders in CQ 1970. 
333 
Laws 
colony has one advantage. The Republic was faced with the problem of 
what to do with adult citizens hardened in un-Platonic ways, and the 
simple remedy proposed was to send everyone over ten 'into the 
country' and educate the children away from the bad influence of their 
elders E406-41 a). Here, although Plato discusses dispassionately 
various forms of purge (????????), with a preference for the more  
drastic,1 he notes with relief that none will be necessary in the present case, 
only a rigorous examination of the candidates for admission. The risk of 
error will always remain, but an adequate period of trial and efforts at 
peaceful conversion should ensure rejection of the unworthy G36b-c). 
The detailed nature of the legislation itself, and the fact that many of 
the institutions and laws copy or refine on those already in force at 
Athens or Sparta,2 also point to a serious practical intent. We are far 
from the disdain expressed in the Republic for those who bother with 
laws about contracts, slander, assault, taxation and so on, which are 
unnecessary in a well-run state of properly educated citizens and 
ineffective in a bad one D25 c-e, 427a). Both works insist that one 
should not try to legislate for everything—some things are better left to 
tradition and public opinion3—and the Laws G88 a-b; cf. 773 c) warns 
against making unenforceable laws which bring the whole system into 
disrepute, but at the same time seems to leave little outside the network 
of legal regulation.4 
Certain phrases have sometimes been taken to indicate that Plato 
recognized his scheme as, like the Republic, visionary. Such are 632 c, 
' Let us refresh ourselves on the way with conversation', in conjunction 
1 Cf. Pol. 293 d, 308-9, p. 185 above. 
* Mainly Athens, in spite of the adoption of ???????? and certain other Spartan institutions. 
Morrow's book shows this in detail (see especially his pp. 232, 271 f., 295, 534k), and cf. Grote, 
PL in, 427f. Jowett (iv, 15) gives a summary of Athenian and Spartan features. Yet one cannot 
read Xenophon's Spartan Constitution or Plutarch's Lycurgus without feeling that in spirit if not 
always in positive enactments, and in spite of his criticisms, P. found much to admire in the 
Lycurgan laws and their aims. See also Levinson, Defense 513-19. 
3 For the importance of ?????? ?????? see 793 a-d. They ought, P. thinks, to be put in writing 
as a code of approved behaviour, though without legal sanction. G93d, and cf. 822e8-23ai. 
The Greek conception of unwritten laws in general is discussed in vol. in, 118-31.) 
4 Private life—'how the individual spends his day'—must be regulated no less than public 
G80a). Cairns in Friedlander {PL 1, 292) gives an impressive list of legally controlled activities, 
including marriage, procreation, distribution of wealth, price-fixing, shipping, merchandising, 
retail trade, innkeeping, the regulation of mines, loans and usury, farming, herding and  
beekeeping, appointment of magistrates, and funerals. I have omitted a few items, nor is Cairns 
attempting completeness. On the regulation of private life see also p. 350 below. 
334 
The city of the Laws 
with 685 a, 'Let us relieve the tedium of our journey by playing an old 
man's sober game of legislation.' At 712, having stated the best  
conditions under which to launch a new state, the Athenian continues to 
Clinias, 'Let us suppose this fiction applies to your city, and like  
grownup children invent its laws in our talk.' It is emphasized that they are 
under no obligation to legislate, and can take as much time as they want 
(857C-58C).1 All this however does no more than reflect the dramatic 
situation. Our travellers are not sitting on a legislative commission but 
walking and talking in the mountains. No one can read the mass of 
detailed legislation proposed (and this must be emphasized here, where 
the omission of much of it might leave an unbalanced picture) without 
concluding that Plato is in deadly earnest about both the purpose and 
the content of his laws. The most reasonable conclusion is that he hoped 
to leave the Laws as a posthumous guide to members of the Academy 
in their business of legislation and to any rulers, such as Hermias of 
Atarneus, who were willing to listen. 
Status and function of laws: the lawgiver as educator. Plato once taught 
that the philosopher, the man of natural wisdom perfected by a Platonic 
education culminating in mathematics and dialectic, should rule  
autonomously, unhampered by laws. In the Politicus we saw this still defended 
as the ideal, but replaced as a practical possibility by the admittedly 
second-best rule of law (pp. i8of., 183-8 above). The Laws takes the 
same position. 
If ever by the grace of God a man endowed with a natural character equal to 
the test could take over the reins of power, he would need no laws to be his 
masters. No law or ordinance is superior to knowledge, nor is it right that 
wisdom should be a slave or subject. Natural wisdom, genuine, true and free, 
should be ruler of all. As things are, however, it is not to be found anywhere 
or anyhow, to any significant extent. So we must choose the second-best, 
ordinance and law, though they can only pay regard to generalities, not to 
every case.2 
Today it is a matter of controversy whether law should concern itself 
with the positive inculcation of morality. Plato had no doubts: every- 
1 It is said more than once that at present their legislation is ???? not Ipyco G36 b, 778 b). 
3 Laws 875 c-d. With the last sentence cf. Pol. 2946-953 (p. 186 above) on law as a blunt 
instrument. 
335 
Laws 
thing to do with laws has a single aim in view, and the right name for it 
is virtue (963a). 'I liked the way you [Clinias the Cretan] embarked on 
explaining your laws. It was right to begin with virtue, and to say it was 
for the sake of virtue that your legislator laid down his laws' F31a).1 
Law is in fact a form of education. 'Anyone who treats of law as we do 
is not laying down the law to the citizens but educating them* (857c). 
'I should like them to be as readily persuaded as possible, and this is 
clearly what the legislator will aim at in all his lawmaking' G18c). To 
this end he proposes to introduce something for which he claims  
complete originality G22d-e), the attachment of prefaces or 'preludes' 
(proemsJ both to the code as a whole (and the Athenian claims that all 
this discussion so far, near the end of the fourth book, has amounted to 
a general preamble3) and to each separate enactment save the most 
trivial. It is tyrannical to impose law 'neat' (??????? 723a), simply  
decreeing that this or that must or must not be done and fixing a penalty 
for disobedience, without an explanatory preface to enlist the  
cooperation of the citizens. The law must mix persuasion with  
compulsion Gi8b).4 To illustrate his point he immediately suggests an  
exhortation to be prefaced to a law imposing an annual fine on bachelordom 
after the age of thirty-five. 
The role of punishment. So far it would seem that, in securing obedience 
to the laws, the consent of the governed is for Plato of primary  
importance. At 690 c, after what sounds a particularly authoritarian passage 
about the accepted right of parents to govern their children,^ well-born 
1 For the contrasting view in Greece see vol. hi, i39f. P. was not in the line of Lycophron 
and Hippodamus, nor of J. S. Mill, Macaulay or most modern opinion, but might have found 
Lords Simonds and Devlin sympathetic. Cf. also Hall, Indiana Law J. 1956, 202, with his reference 
to 'the thoroughgoing social character of Plato's ethics'. 
2 In music a prooimion was a prelude to the main theme, and P. plays on the double meaning 
of nomosy law and tune. See 722 d. According to Pfister in Melanges Boisacq 173-9, Plato's 
prooimia were not so original as he claimed. 3 Cf. also the'address to the colonists' at 715 eff. 
4 This seems a praiseworthy idea, but is harshly treated by Versonyi in his article on Morrow's 
book (?. of Metaph. 15, 1961-2, 69f.). Why he thinks that P.'s preface contains no 'rational 
instruction' I do not know. It is not incompatible with persuasion, and the doctor of 720d, whom 
he mentions, ???????? the patient, as good doctors do. 
For a lawyer's opinion see Hall, Indiana Law J. 1956, 182 n. 52: 'The preambles in Laws 
include A) references to the principles which supply the rational basis of the enactment and 
B) exhortation to obey the law.' 
5 One would not expect P. to support the saying' maxima debetur puero reverentia', especially 
in view of his characterization of children at 808 d as, owing to their still uncanalized powers of 
336 
The city of the Laws 
the lower orders, elders their juniors, masters slaves, even of the stronger 
to rule the weaker,1 he concludes that the strongest claim of all is that of 
the wise to rule the ignorant, and that this is achieved most naturally 
by the rule of law over willing subjects without force. But this again is 
not the last word. The lawgiver must not curry favour with the  
populace. To legislate for the people's pleasure would be like expecting 
medical treatment or hard training to be pleasant in itself F84 c). This 
leads naturally to the role of punishment, though even to consider it, 
says the Athenian at the beginning of bk 9, is a kind of admission of 
failure in a city supposed to be founded on the right lines and to give 
every incentive to the practice of virtue. Laws are made for good men, 
to assist them to live amicably together, but some must be for those 
who spurn such instruction (88od-e). These the lawgiver hopes he will 
not have to use, but unfortunately human beings are not perfectible, 
and unteachable persons are bound to appear. In the event we find that 
every law carries its appropriate penalty, ranging from a vague '  
reprimand ' or ' loss of reputation'2 through fines, loss of civil rights and 
exile, to death. On the one hand the aim is therapeutic, and in a 
remarkable passage (862d-63e) Plato says that it may be achieved not 
only by punishment but by talking with the offender and even offering 
him pleasure, honours and gifts. Any means is right that will heal the 
criminal's diseased mind and bring him to hate injustice.4 Only if he be 
reason, 'Hardest to manage of all wild things . . . cunning, shrewd and insolent'. Yet Juvenal 
might almost have been translating his ??????????? toOs ????*. Nowadays, he says G29 b), 
admonitions to the young to be respectful are ineffective. ? wise legislator would rather bid 
adults respect the young, and above all beware of letting their juniors see or hear them doing or 
saying anything disgraceful. Where the old have no shame the young will show no 
respect.' 
1 This may be a shock to those who remember their Gorgias. Clinias's rejoinder—'There's 
no getting away from that'—shows that he takes the words in their commonest sense, not as 
'better' and 'worse'. One has to remember that P. has Pindar in mind F90b8), and take the 
passage with 714?, where he points out that some of the claims listed here are mutually  
incompatible, and 890a. (See England on 690b8.) The passage of Pindar has been discussed in vol. in, 
131-4· 
2 Certain unspecified honours and awards for merit (????????) are mentioned, from which an 
offender will be disqualified (845 d, 935 c, 952d). For the awards see also 961 a, and Morrow 271 
n. 65. Public opinion is often invoked against an offender, e.g. 762c, 880a, 914a, 917c, 
936b. 
3 A good example of psychiatric talk is the preface to the law on sacrilege, 854b-c. 
4 Naturally this demands close attention to the circumstances of the crime and the criminal's 
state of mind at the time. Saunders (CQ 1973, 235; PQ 1973, esp. p. 353) refers to passages 
showing that P. was fully aware of this commitment. The conception of wickedness as a mental 
23 337 ghg 
Laws 
judged incurable, then for his own sake (for life is no boon to such 
people, 862e) as well as the community's he must be put to death. In 
practice Plato by our standards make pretty free with the death penalty. 
It is exacted for instance not only in certain cases of deliberate murder1 
but for sedition (854b-c), open atheism (at the second offence, 909 a), 
temple-robbery by a citizen (854c), persistent perjury in court (937c), 
acceptance of bribes when in office (95 5 d), perversion of justice from 
motives of greed (938c), and disseminating harmful notions from 
abroad (9520-d).2 For certain crimes a citizen must be put to death 
but not a slave or foreigner, because the citizen has had the right 
upbringing and yet proved incurable. The others may still be brought 
to see reason, and to this end they will merely be branded, given as 
many lashes as the judges impose, and cast naked over the frontier. 
They might well agree with Plato that death is preferable !3 
Imprisonment as a punishment or corrective4 is occasionally resorted 
to, notably for atheism. As he explains in bk 10, Plato considers  
disbelief in the divine and rational governance of the world to be the root 
of most moral evil. Open atheists of otherwise blameless life are sent for 
at least five years to a kind of mental home (??????????????) where 
they will be visited only by members of the Nocturnal Council (pp. 
309ff. below), who will admonish them' for the salvation of their souls \5 
disorder raises the problem of how to reconcile punishment with the Socratic teaching that 
wrongdoing is due to ignorance and no one is voluntarily wicked. To this dilemma P. turns his 
attention in bk 9, and though it might be thought appropriate in the present context, I have 
deferred it to a final section on some philosophical points (pp. 376-78). 
1 The laws on homicide are complicated, as at Athens, by considerations of religious pollution 
(871 a-d, 873b; cf. Grote, PL in, 404f.), and also by P.'s belief in posthumous punishment and 
reincarnation; e.g. a matricide will be reborn as a female (872c). In some cases where today 
capital punishment or life-imprisonment might be though suitable, purificatory rites are deemed 
sufficient, in others (murder of kin) purification plus a period of exile. 
2 The death penalty, at least for some offences, is to depend on a majority vote of the dicasts 
(856c), as at Athens in the trial of Socrates. 
3 But at 938 c the foreigner will suffer only banishment for an offence for which a citizen must 
be put to death. Penology may well have been a subject which P. intended to revise. 
4 A thief will be detained in prison until he has paid twice the value of the theft unless excused 
by his prosecutor (857a). A citizen who illegally engages in retail trade will be imprisoned for a 
year in the first instance (9196-20 a). See also 880 c. Morrow points out (p. 294) that at Athens 
prison was as a rule only used for holding defendants for trial or the condemned until execution. 
5 909 a ??! ?????????? ?? ??? ??? ????? ???????. The dangers of such a system seem 
obvious today, when sinister analogies suggest themselves, and are pointed out by Morrow D91 f.). 
Saunders however (CQ 1973, 235), though properly critical of P.'s assumptions, perhaps more 
justly calls the interviews 'serious philosophical discussions' and of all P.'s methods of assessing 
338 
The city of the Laws 
Should the 'cure' prove ineffective after release, a second conviction 
carries the death penalty. Those on the other hand who, unbelievers 
themselves, prey on the superstitious fears of others, are  
condemned to solitary confinement for life in a remote part of the 
country.1 
Emphasis is frequently laid on a citizen's duty to lay information if 
he knows a crime has been committed, and also (which sounds more 
dangerous) to carry out summary justice on his own account, inflicting 
physical punishment on anyone, slave or free, whom he sees  
misbehaving. ' He who commits no crime deserves honour, but doubly so he who 
checks the wickedness of others ... by revealing it to the authorities; 
and whoever joins with them in punishing it is the perfect citizen and 
takes the palm for virtue.' Honours and punishments are assigned 
accordingly.2 
To conclude, the severities of the penal code are due to Plato's 
keeping his principles as high as ever but at the same time trying to 
incorporate them in the institutions of a particular earthly state.  
Education, explanatory prefaces and the laws themselves are designed with 
one aim only, that conscious of the underlying reasons for the prescribed 
code of conduct the citizens will voluntarily adhere to it and serious 
transgression will disappear or be confined to a few perverted  
characters. If in spite of these exceptional advantages some do go seriously, 
even irreparably astray, their treatment must be correspondingly 
drastic. We must also remember that death in Plato's eyes was ' the least 
of evils' (854c), for exactly the same reasons as long before in the 
Apology: the good of the soul is all that matters, and for all we know, to 
meet the gods of the next world may be the best thing that could happen 
to us G2yd). 
Theory and reality. As enumerated by himself, Plato's aims could not be 
faulted by any modern democrat. They are to foster, besides good 
sense, liberty, equality, and the spirit of concord—the slogans of the 
character 'certainly the one which approaches most nearly to modern psychological practice'. 
See also vol. in. 246. 
1 When they die they will be denied burial, but it is hard to see why they are not put to death 
at once. No hope of a cure is held out. 
2 73od. For examples of summary punishment by private individuals see 762c, 914b (of 
slaves), c>i7c-d, 935c. 
339 
23-2 
Laws 
French Revolution.1 How do these admirable ideals look when we 
examine his means of attaining them? The prime requisite is the mixed 
(or intermediate, ?????) constitution already referred to (pp. 3 31 f. above). 
neither unchecked dictatorship nor extreme egalitarian democracy. 
These are 'non-constitutions', implying simply the exploitation of one 
section by another. (Cf. ? 14 A.) In his own state the law will ensure that 
no citizens interfere with any others and each has time and opportunity 
to pursue the best life.2 There can be no friendship if rulers and ruled 
live like master and slave (hence liberty and friendship go closely  
together), nor on the other hand if good and worthless are treated alike, 
for 'equality between unequals is inequality' G57a),3 and either extreme 
leads to discord. In what sense then is equality good? We have seen the 
answer in the GorgiasA The word covers two contrasting procedures, 
the easy one of giving equal shares^ to all indiscriminately (arithmetical 
equality) and the less obvious one of distribution in proportion to 
individual merit (geometrical equality). The second is the 'truest and 
best' because it secures justice, and justice is preferable to dictatorship 
and even to the power of the people. 
In Magnesia the ideal of the mixed constitution will be achieved by 
providing that all magistrates, officials and members of boards charged 
with administering the laws will be elected by popular franchise and 
serve for a fixed term of years only,6 after which, as at Athens, they 
will be called to account before a board of examiners, also popularly 
elected. The methods of election, age qualifications and length of 
service will vary with the office, and are prescribed in detail for each. 
1 693 b and d, repeated at 701 d: (????? ????????? ?? ????? ??? ??? §?????? ??? ????? ?????), 
757a5~6> 739^ °f tne ideal state, ??????? ????? ???? ??? ??????? ????? ????????????. The 
importance of ????? is 'repeatedly emphasized' (Morrow 562 with 5 reff.). 
2 832c-d. Cf. 7i2e-i3a. 
3 757a. At Rep. 558c he had described contemporary democracy as 'dealing out a kind of 
equality to equals and unequals alike'. 
4 Vol. iv, 301, where Laws 757b-d was quoted in explanation of the phrase 'geometrical 
equality' at 508a. The conception of geometrical equality, in both its mathematical and its 
political aspects, has been fully investigated by F. D. Harvey in Class, et Med. 1965. For theLaws 
see pp. io8f. 
5 Or chances. The Greek has no noun, and P. is here thinking of office rather than wealth. 
As an example of his meaning he mentions appointment by lot G57b). 
6 As Morrow says A62), we should regard this as truly democratic, but at Athens the  
requirements of democracy and equality seemed only to be satisfied by the use of the lot. This was of 
course entirely opposed to P.'s principle of giving office to the best qualified. Nevertheless he 
does allow for a strictly limited use of sortition, avowedly to avoid popular discontent. (For the 
34O 
The city of the Laws 
No man must be reckoned subject to another, but all alike will 
be subjects of the law. The doctrine attacked in the Republic^ that 
to serve the interests of those in power is justice, is again rejected 
G41 cff.). Laws must benefit the whole state, not a party, and  
officeholders like anyone else are their servants.1 
C) Life in Plato s city 
Population. Before we go further, something must be mentioned which 
is of primary importance whenever Plato's social and political theories 
are compared—as they always have been and will be—with the theory 
and practice of more recent times. That is, the si{e of the community for 
which he is planning. Speaking as we do of democracy, aristocracy, 
oligarchy, tyranny, in terms taken straight from Greece (though to 
make it sound more modern we may replace the last by 'dictatorship'), 
we are apt to think that they originally stood for different forms of 
government in societies roughly similar to our own. We rarely pause to 
reflect on the difference of scale, sufficient to entail a difference in kind. 
We count our populations in millions—over fifty in the case of Britain 
—whereas Plato posits a community of 5,040 households,2 each with 
its separate allotment of land. Morrow has estimated that this would 
mean 10,000-12,000 male citizens and a total citizenship of 40,000 to 
details see Morrow 161-3, 233.) In some things he is indeed learning to come to terms with 
practical politics. (On the examination after laying down office see Morrow 220-7. Itwas especially 
directed against corruption.) 
1 ?????????, ?????? ??? ?????. See ji$Si-a, This rule of law, to replace irresponsible use of 
power, whether by a tyrant or the mob, is violently attacked by Versonyi in R. ofMetaph. 1971-2, 
77f.—curiously so for a scholar whose work is in general a model of restraint and fairness, as are 
the last two pages of the same article. As will appear, I cannot accept his statement that' in Plato's 
Cretan state there will be no philosophers or philosophy'. Philosophy will of course be confined 
to a small number, as in Rep. and everywhere else. 
2 737e~38a. It is typical of the difference between Rep. and Laws that in the former the state 
is simply to be 'of a size to which it can grow without losing its unity' D23b), with no hint of 
what that size might be, or how to ensure it. The expressed reasons for the choice in the Laws are 
purely practical, first generally (self-sufficiency, defence, 737 c-d), then, in relation to the precise 
number 5,040 (??????? ???????????? 738a), the presence of a large number of consecutive 
divisors, useful for organization in war and peace, for contracts, taxation and distributions. 
Contrary to general opinion (cf. Taylor, PMW *?? n. 1, and Morrow 428) Kahn sees a deeper 
significance in it as an imitation of unity, and connects it with making the state' as near immortality 
as possible'. (Cf. 739? and Kahn in//// 1961, 422.) Bardies in ?????? 1971 thinks of the special 
status of the number 7. (? ?2?3?4?5?6?7 = 5040.) But we may be content with what 
P. tells us. 
341 
Laws 
48,00?.1 This is roughly the population of Famborough or half that of 
Bath, distributed of course between city and countryside. In addition 
there would be 7,000-8,000 metics and perhaps 30,000 slaves. In a state 
of this size, as at Athens, important functions, including the election of 
officials and trials for offences against the state, could and would be 
performed by the whole body of adult citizens meeting in Assembly.2 
Public before private weal. ' The law is not concerned to promote the 
especial welfare of a single class, but to ensure that of the city as a 
whole' {Rep. 519c; cf. 420b). This precedence of common interests 
over sectional or private is maintained in the Laws, even if as an ideal 
which admittedly can never be realized to perfection. 
The first and best city, constitution and laws exist where the old saying applies 
universally, that the property of friends is truly common to all... where by 
every means the word * private' is eradicated from life and things are as far as 
possible so contrived that even what nature has made one's own ,like eyes, 
ears and hands, seem somehow to see, hear and act for the community. All 
praise and blame as one, and feel pleasure and pain at the same things. No 
one could propose better evidence of superlatively good laws than that they 
give to a city the greatest possible unity. 
To this end both the laws and the educational system (to which in 
Plato's mind the laws belong) are principally directed. To take one or 
two illustrations, although the citizens will have their own plots of land, 
because farming in common would be too great a strain on men born 
and bred as they are now, each must regard his own plot as the common 
property of the state, cherishing his native land as a child its mother, 
1 See his PCC n8f., 129 n. 105, for these figures and comparison with the population of  
existing Greek city-states. According to Critias A12d), in the mythical and ideal primitive Athens 
care was taken to keep the population of military age, men and women, at about 20,000. The 
question how to maintain the population at the recommended size is not neglected. Each  
householder will have one heir and give his other sons for adoption to the childless. The birth-rate 
can be stimulated or reduced by educating public opinion through marks of appioval and  
disapproval, and in the last resort excess of population can be checked by the old Greek device of 
colonization. See 74ob-4ia. 
2 Women can hold office (????? 785 b), so a fortiori would be members of the Assembly. 
Aristotle denned a citizen as one who is eligible for deliberative and judicial office {Pol. 1275 b 18). 
Cf. Laws 768b: 'Anyone debarred from taking part with others in judicial proceedings feels 
himself altogether excluded from citizenship' (Morrow 128). 
342 
Life in Plato's city 
indeed as a divine being.1 Again at 923, in the preamble to testamentary 
laws, the citizens are told that they and their property belong not to 
themselves but to their families—ancestors, relations and descendants— 
and in turn the whole family and its property being to the community 
(923 a-b). Even in marrying, the partners must consider the  
community's best interests, not their own preference. Here the point from 
the Politicus is repeated, that a union of opposite temperaments  
maintains the best balance of character, though it is natural for like to be 
attracted to like.2 To put public good before private in this way is in the 
end best for the individual too, who cannot escape being a part of 
society, nor indeed of the whole cosmic scheme. In bk 10 the  
thoughtless youth who does not believe that God can care about the behaviour 
of a single individual is told that like a good craftsman God attends to 
details as well as the general plan. All have something to contribute. 
' Creation is not for your benefit; you exist for the sake of the whole ... 
You grumble because you don't see that as far as you are concerned, 
what is best for the Universe is best for you, by virtue of your common 
origin. '3 
Private property: the four classes. In the Republic D2ie-22a) Plato was 
determined to keep riches and poverty out of his state, but gave no hint 
of how he would do it, save by forbidding the military and governing 
classes to have any possessions at all. Here again, the Laws reafBrms the 
goal G286-293, 744 d) and fills in the practical details. The principle to 
be observed is that * money must come last in the scale of values', after 
the proper care of soul and body G43 e). All will live directly off the 
land. Gold and silver will be replaced by a token currency,4 and the 
common coin of Greece reserved for military expeditions and officially 
1 740a. This of course is an appeal to popular religious belief in the Earth-Mother. Cf. Tim. 
4ob-c, and for the Athenian claim to be literally offspring of Attica, Menex. 237 d-e, vol. iv, 
315 n. 3,463. 
2 733a-€· Cf. Pol. 3iob-i 1 a, and p. 190 above. P. admits that legal compulsion here would be 
unpopular and ridiculous. Only by explanation, exhortation and reproach can people be  
persuaded of the importance of producing well-balanced children. Marriage itself however, by a 
certain age, is compulsory. 
3 Laws 903b-d. The 'common origin' has of course been explained in 77m. 
4 Another change from Rep., where the possession of gold and silver will only be denied to 
the guardian classes D17a). P. is following Sparta's example. See Xen. Rep. Lac. 7.5, and for the 
iron currency of Sparta [Plato] Eryxlas 400a, Plut. Lye. 9. 
343 
Laws 
authorized travel G4ie~42c). Though it cannot be arranged that each 
citizen arrives with equal means G44b), discrepancies of fortune will be 
strictly limited. Each citizen will be granted a holding of equal value,1 
which may not be sold or otherwise disposed of outside the family. Thus 
the poorest will have sufficient from their farms to support family and 
slaves, and at the other end of the scale no one may possess wealth 
exceeding four times the value of the plot.2 (Wealth must be declared 
and registered, and any surplus handed over to the state, 7446-453.) 
For administrative purposes, and in accordance with the principle that 
'the truest equality is inequality in proper proportion', there will be 
four property-classes, and a man's class will make some, but not much, 
difference to his status and prospects. (There will of course be transfers 
as fortunes vary.) 
In the filling of offices these property-classes play a very minor part. They 
are disregarded completely in the selection of the most important officers of 
the state, viz. the guardians, the euthynoi, the educator [' Minister of  
Education'; see below, p. 346] and the members of the court of select judges . . . 
also . . . generals and other military officers. All citizens are admitted 
to the assembly and to the popular law courts without consideration of 
property.3 
The Council·* will consist of an equal number from all classes, and in its 
election all citizens have the vote. However, at certain stages of the 
complicated five-day procedure, members of the lowest class, or the two 
lowest, are excused the fine imposed for not voting G56b-e). By  
provisions such as these Plato hopes to minimize any sense of grievance or 
unfairness between his citizens. 
1 By varying the size according to the quality of the soil. Further details of the elaborate 
pattern of allotment will be found at 745 b-e, e.g. each family is to have two homes, one in or 
near the city and the other near the frontier. In general, and for the historical precedents, see 
Morrow 103-12. 
2 There are four property-classes, but perhaps it is more accurate to say with Aristotle that the 
highest owns^ve times the minimum. See Morrow, A. and P. in Mid-Fourth C. 146?. 
3 Morrow 133^ Property qualifications are required for temple-treasurers and ????????? 
(city-wardens, Class I only), ????????? (country-wardens, I and II) and supervisors of athletic 
contests (I-III). For further details, and comparison with Greek practice, see Morrow 131-8. 
Public office carries no salary (Morrow 191). 
4 I.e. the Boule, corresponding to that at Athens and not to be confused with the Nocturnal 
Council, which is called a ???????? (95id, 961a et ??.). 
344 
Life in Plato s city 
Trade and labour. No citizen may engage in trade or ply any craft. 
These must be left entirely in the hands of resident foreigners, not as 
dishonourable but because no human being can carry on two callings 
efficiently, and a citizen's occupation is to maintain the social order 
(kosmos) which he enjoys, a task demanding his whole attention, with 
much study and practice.2 In this the whole citizen body corresponds to 
the guardian classes in the Republic^ but without being deprived of 
personal property and family life. For the same reason, though a  
householder will manage his farm and supervise the work, the actual labour 
will be undertaken by slaves.3 At 806 d-e Plato, raising the question of 
how his citizens will spend their time, notes that their basic needs have 
been provided for, manufacture has been left to others, their farms are 
worked by slaves who provide them with enough produce for men of 
temperate habits, and there are the common meals both for men and for 
women and children.4 
Education.** In spite of its supreme importance for Plato, the subject of 
education is treated in a more rambling way, and with more irrelevant 
1 So it was in Lycurgan Sparta. See Xen. Rep. Lac. 7.2, Plut. Inst. Lac. 239d. 
a 846c!. It is the principle of'doing one's own', declared in Rep. to be the essence of Justice. 
For the law relating to trade see 919 d, 920a, and cf. Morrow 141-6, especially for the role of 
metics in historical Greek states. 
3 P.'s intentions are not altogether clear, ?????????? and ??????????? (8o6d-e) would  
normally refer to letting and paying rent, and are so taken by Taylor ('let out to villeins', trans, 
p. 101) and LSJ s.v. ????????. The anachronistic 'villein' gives one pause, but P. may have in 
mind a status like that of the helots at Sparta, as Morrow and others have thought. See Morrow, 
PCC 149 (but contrast 150 and 151!) and A. and P. in Mid-Fourth C. 152. 'The helots worked 
the land for them', says Plutarch, 'paying the appointed tribute.' (See Plut. Lye. 24 and other 
reff. in RE viii, 205.) Morrow adds, however (opp. citt. 531 and 152), that in the farming laws of 
842eff. 'Plato clearly implies that his citizens will be tillers of the soil'. I would say these laws 
imply no more than that they exercise supervision and give orders. At Pol. 1265 a7 Aristotle says 
that they are free from all menial (????????) tasks, resembling in this the guardians of the 
Republic; and these, as he has just said, do not do their own farming but are provided for by the 
third class A26439 and 33; cf. Rep. 416d-e, Critias noc-e). 
4 The Spartan institution of common messes for men is extended to women for reasons given 
at 781 aff., but family life is not abolished as in Rep.: after the meal and due libations 'they all go 
home' (807a). How the ???????? are provided for is uncertain. The threefold division of  
household produce, for free citizens, slaves, and craftsmen or other foreigners in residence (8476-48 a), 
leaves none earmarked as a contribution to the common table. See however Morrow 395 f., and 
for the Spartan system of individual contributions Plut. Lye. 12. 
5 For full details of the educational curriculum see Morrow ch. vn. Grote, PI. in, 376-85, is 
also worth looking at. Bury's' Theory of Education in Plato's "Laws'" describes P.'s educational 
aims in somewhat lyrical terms, with timely remarks on the relationship between ??????? and 
??????. 
345 
Laws 
digressions, than any other major topic. Its purpose is defined more 
than once. In bk ? F43 e) it is ' that cultivation of excellence {arete) 
which fills a child with eager desire to become a perfect citizen, knowing 
both how to rule and how to submit to rule with justice'. True  
education is contrasted with vocational or other training which is banausic, 
illiberal and unworthy of the name. Bks 1 and 2 showed the educational 
importance of play, song and dance. This is resumed in bks 6 and 7,1 
where also a more academic curriculum is mapped out. There will be a 
Director of Education, with assistants appointed by himself (813 c), who 
must be considered supreme among the highest ministers of state.2 
Children from three to six years old attend village play-groups, at 
which, under supervision, they devise their own games. Schooling 
begins at six, and is compulsory,3 because 'children belong to the  
community rather than to their parents'.4 The teachers will be foreigners, 
paid by the state.5 Boys and girls should have exactly the same  
education (including physical and military training, 804 d-e), though  
segregated in different schools G94 c-d). At first the chief part will be played 
by dancing and singing, about which bk 7 has a great deal more to say. 
Reading and writing will be learned from ten to thirteen, music  
(especially lyre-playing) from thirteen to sixteen. Basic literacy is essential, 
but high speed and calligraphy should not be demanded of slow learners 
(810b). At this stage the Director of Education will need to supply 
reading matter. Not all Greek literature is suitable, and Plato thinks his 
best guide will be the Laws itself.6 Teachers should be told to study this 
1 What P. says about the psychology of Corybantic and Dionysiac frenzy, and its ritual 
indulgence as a homeopathic cure for irrational fears, is remarkable. See 79od-<)i b with notes in 
Saunders, Penguin trans. 274, and Grote, PL in, 376-8. 
2 765 c For this office, its duties and mode of election, see Morrow 324-6. 
3 794c-d. 'As far as possible', adds P.—further evidence that in the Laws, as nowhere else, 
he is conscious that politics is 'the art of the possible'. 
4 Sparta again. Cf. Plut. Lye. 15: 'Lycurgus considered children not as belonging to their 
fathers but as the common property of the state.' 
5 804 c-d. The need for payment is probably, as Taylor says (PMW 484), the reason why 
citizens are not employed. Morrow C76 n. 102) remarks on the low esteem in which teaching was 
held in Athens as a profession for citizens. It is nevertheless surprising that P., for whom education 
meant above all inspiring the young with his own ideals of citizenship, should have entrusted 
them to foreigners in these impressionable years. 
6 'The discussions which we have been having from dawn until now' (811 e). These of course 
have not been written down, but on the side of dramatic plausibility it may be said that Clinias 
as a founder would certainly carry away a vivid impression of them, and the Athenian would no 
doubt help, as he will in the programme of higher education (968 b). 
346 
Life in Plato's city 
and teach it to their pupils, together with such poems or prose-works 
as are in line with its precepts. 
Other subjects to be taught (at what age is not specified) are  
arithmetic, mensuration and astronomy, not to a very advanced stage and 
with largely practical intent: letters and calculation for running a  
household, a state or a war, astronomy in order to understand the grouping of 
days into months and years for the organization of festivals and other 
honours to the gods (809c-d, 819c). Numbers are best taught from 
earliest years through play, sharing out apples between the children and 
so forth.1 One thing they must learn. Ignorance of it is shocking, 
though even the Athenian (presumably Plato) only learned of it late in 
life and blushes for his former 'swinish' stupidity. This turns out to be 
the existence of incommensurables,2 which he claims is not difficult to 
understand and may also be taught in play. Astronomical teaching too 
has its scandal to be removed, the heresy that the motions of the divine 
planets are irregular. This again the Athenian has only escaped at a 
mature age, but if he can prove the point it must find a place in the 
education of the young to save them from blasphemy. Instead of doing 
so however, he turns at this point from education to the laws on  
hunting. It would, after all, only repeat the Timaeus, and perhaps we may 
suppose that Clinias and Megillus would scarcely be capable at their 
age of taking it in. 
Slavery.^ In the Republic Plato says so little about slavery that some 
have thought he favoured its abolition. This is impossible,4 and in the 
Laws he has given much thought to this firmly entrenched Greek 
institution. Slaves will be both public^ and private. Ownership of 
slaves, he says G76b ff.), teems with difficulties. Slaves have been known 
to prove better than brother or son, the salvation of their master's 
1 Grote's note on method here {PL in, 383 f., n. ?), though a century old, is still of interest. 
2 For the importance attached by P. to the discovery of irrationals or incommensurables see 
Popper, OS 1, ch. 9 n. 6, pp. 248-53, already referred to in connexion with 77m., p. 283 n. 2 
above. 
3 This is an exceedingly complex subject. Besides Morrow's PCC see his * P.'s Law of Slavery 
in Relation to Greek Law' and Gernet, Laws (Bude" ed.) 1, cxix-cxxxii; on Greek slavery in 
general, Slavery in Class. Ant. (ed. Finley) includes a 'Bibliographical Essay'. 
4 See vol. iv, 483 n. 1. 
5 Tfjs ?????? ??????? 794 b. Privately-owned slaves may be requisitioned for public works, 
as far as possible when not required by their owners G60 c—61 a). 
347 
Laws 
person, household and property. Yet some dismiss the whole class as 
depraved and untrustworthy and treat them like animals, making their 
souls a hundred times more slavish than before. The whole distinction 
between free man and slave is not easily maintained; witness the  
frequent slave revolts. The best policy is, first, to have slaves as far as 
possible of different nationalities and languages to make combination 
difficult,1 and secondly to train them well both for their own sakes and 
one's own. For this the primary requisite is to treat them as humanely 
and justly as one would an equal, as indeed anyone in authority ought 
to behave to those weaker than himself, thus sowing the seeds of  
goodness in them by example. One should none the less be firm, punishing 
where punishment is merited (a rule which applies no less to one's own 
children, 7933-943, 808e), and avoid familiarities, which only make life 
more difficult for slave as well as master. 
The legal proposals sometimes shed a different light. Here are a few. 
Slaves may be freed, but retain fairly onerous obligations towards their 
former masters (915) and must leave the state after twenty years (i.e. 
are treated as metics). A slave (or foreigner) convicted of stealing public 
property will be fined (evidently slaves will possess money) or  
otherwise sentenced by the court on the assumption that he is probably 
curable, whereas a citizen who in spite of his upbringing commits 
robbery or violence against his country must be put to death as 
incurable (941 d~42a). Murder of an innocent slave to conceal one's own 
crimes carries the same penalty as murder of a citizen (872 c). A slave 
who murders a citizen will be delivered to the victim's family for  
execution in what manner they will.2 In certain cases (e.g. neglect or ill- 
treatment of parents) a slave will be freed for laying information and 
protected against retaliation.3 Severe corporal punishment may be 
inflicted on a slave, e.g. for striking a citizen (882a-b), though a slave 
may be employed by the magistrates to mete out similar punishment to 
a free criminal (882 a-b). Some proposals are both bizarre and brutal. 
1 No doubt P. would still forbid the enslavement of Greeks captured in war, as he did in 
Rep. D69 b-c). 
* 868 b-c. That is, if the slave was carried away by passion. If the killing was cool and  
deliberate, he is to be taken by the public executioner within sight of his victim's tomb, scourged there, 
and if he survives the scourging, executed. 
3 932d. In another case, his failure to inform carries the death-penalty (914a). 
348 
Life in Plato's city 
If a man strike his parents or grandparents, bystanders are bound to 
come to their aid. A slave who does so earns no less than his freedom, 
but if he does not, will receive one hundred lashes.1 Such unfilial 
assaults, one imagines, are unlikely to occur often in public places, but 
Plato's language shows that the horror with which they inspire him is 
to a large extent religious. From 794 a-b it appears that infant slaves will 
share in the state-organized village play-groups for children between 
the ages of three and six,2 though not in the later stages of state-educa- 
cion. There are laws to prevent the sale of unhealthy or otherwise 
inferior slaves (916a-b) and for dealing with the offspring of liaisons 
between slave and free (93od-e). 
The whole conception of slavery has become so foreign to us that I 
prefer to give some information and leave moral judgements to the 
reader. He may care to consider Kahn's {JHI 1961, 424): 'There can 
scarcely be any doubt of Plato's natural humaneness: this is evident... 
even in his general remarks on the treatment of slaves (vi, 777d). Yet 
his humane sentiments are so utterly overruled by his sense for order 
and hierarchy that he proposes a slave legislation harsher and more 
retrograde than that of his own time.' 
Daily life in Magnesia. Relieved of life's drudgery, are his citizens, asks 
Plato rhetorically (807 b), to live like fattened cattle, at the mercy of any 
lean and spirited beast that comes along? Even the second-best state, 
lacking the perfect communism of the Republic, can do better than that. 
Since however a mass of trivial legislation on the daily round would be 
unseemly, he confines himself to a moral homily. Every day and night 
would hardly suffice for the mental, ethical and physical training  
demanded by citizenship, so from one dawn to the next a free man's life 
should run to a time-table. Body and soul need far less sleep than is 
normally taken, and much of the night should be spent on one's own or 
the state's business. Officials will be abroad in the city, feared by the 
wicked and admired by the just. The master of the house should be the 
first to be up and about, and the mistress should call the maids, not vice 
1 881 c. Twice the legal maximum at Athens (Morrow, 'Law of Slavery' 69, Gernet, Bude" 
introd. cxxv n. 2). 
2 This may be a misinterpretation, though understood in the same way by Morrow (P.'j Law 
of Slavery 44). See Saunders in CR 1961, 101. 
349 
Laws 
versa. The days will be occupied either in public or legal business (for 
every citizen either is, or must be prepared to be, in office or on a panel 
of justice) or in healthy and pious enjoyment of the festivals dedicated 
to the gods. These are numerous,1 and might be described as a sort of 
'Kraft durch Freude' institution, offering at the same time a rest from 
duties and an opportunity for self-improvement F53 d). Like 
all Greek festivals they included athletic contests as well as dancing 
and singing, the moral importance of which we have already noted. 
Second to divine favour, their organization has a definite social aim, 'to 
encourage mutual acquaintance and social contact of every kind' 
G71 d). 
The above suggests that Plato would advocate more interference in 
private life than we should regard as tolerable,2 and in fact he lays it 
down as a principle that 'without the proper regulation of private life it 
is vain to expect any firm foundation for the laws on public affairs' 
G90b). To make it a matter of legal compulsion, he continues, would 
be to invite ridicule and a flouting of the law, but the rules should  
nevertheless be spelt out without attaching any penalties, in the expectation 
that free men will treat them as laws and experience the happiness of 
having well-managed homes and city.3 If moreover the state seems to 
intrude excessively into the private life of the citizen, he will at least be 
protected from annoyance by his neighbours. The non-interference 
of individuals in each other's lives will be secured by the provision for 
private suits in the courts G68 b-c), the composition and procedure of 
which are laid down in elaborate detail and to a large extent follow 
Athenian practice. The brief preamble to the penalties for cultivating or 
grazing cattle on a neighbour's land, or enticing his bees away, is worth 
quoting for its human touch.4 It runs (843 b-c): 
Next come the many petty injuries among neighbours. Their frequent 
repetition leads to considerable hostility and thoroughly embitters neigh- 
1 'Everyone partakes in a lifelong round of sacrifices, festivals and choric song and dance' 
(835 ?). See Morrow 353 f., and on the whole subject of festivals 352-89. 
2 Cf. p. 334 n. 4 above. 
3 780 a on the other hand does advocate the control of private life by actual legislation (in 
leading up to compulsory attendance at the common meals). 
4 For Plato, it would seem, de minimis curat lex. At 925 dff., after some stringent marriage- 
laws, he adds that they will bear heavily on some people and the law must allow for hard cases. 
' Some may think that these are not the legislator's concern, but they are wrong.' 
350 
Life in Plato's city 
hourly relations. Therefore everyone must take the greatest care not to offend 
his neighbour, especially in the matter of encroachment on another's land. 
It is easy to injure a neighbour—anyone can do it—but not everyone has the 
chance of doing him a favour. 
Finally, something must be said about a passage on discipline and 
command, which has suffered from one-sided interpretations.1 In bk 12, 
942 a-d, Plato returns to the subject of military service (???????? 
942a 5). For this it is essential, he says, that no man or woman be 
without a superior (anarchos), whom he or she will obey in every detail, 
never acting independently. The soldier must be so habituated to acting 
as one of a group that the thought of doing anything on his own never 
enters his head, for this is the best and most efficient recipe for survival 
and victory in war. We recognize here the principle of discipline and 
the chain of command familiar in armies of our own day. In peacetime, 
he continues, from childhood upwards, we must practise this art of 
ruling others and being ruled by them in turn. The absence of a ruler 
(anarchia) must be wholly eliminated from the life of everyone. 
Since the Second World War,2 this passage has been branded as an 
outrageous example of totalitarian ethic. 'Like other totalitarian 
militarists and admirers of Sparta, Plato urges that the all-important 
requirements of military discipline must be paramount, even in peace, 
and that they must determine the whole life of all citizens', who must 
'spend their whole life in a state of permanent and total mobilization' 
(Popper, OS 1, io3).3 Since this would be wholly opposed to Plato's 
recommendations elsewhere in the Laws, the passage must be looked 
at more closely. First, the total submission to orders is introduced as a 
necessity for military training and active service. The things to be done 
at command are enumerated: standing still, marching, physical training, 
washing, eating, night-duty as sentry or despatch-carrier, and in actual 
warfare pursuit of the enemy or retreat. Plato's love of analogy can 
mislead. He compares the legislator or archbn here to an army officer 
1 Dr Saunders when he read this chapter drew my attention to M. J. Silverthorne's article 
' Militarism in the Laws}' in Symb. OsL 1973. It does not conflict with what is said here, but makes 
some different points, and I think both accounts may be left to stand independently. 
3 And not only since then. Gomperz has some exaggerated and partial comments in GT vol. 
in, 262. 
3 See also Popper's 'Reply to a Critic' in later editions, pp. 338-42. 
351 
Laws 
just as he compares him elsewhere to a doctor. Clearly it is army life in 
camp or battle that is being described. It has however, Plato believes, 
an analogy in ordinary life, part of which will in any case be spent in 
military training.1 The analogy to indiscipline in the forces is anarchy2 
in civil life. There however the ideal is not for everyone to suppress 
initiative and blindly follow the orders of a superior. The art for a 
citizen to cultivate is 'how to rule and be ruled in turn' F43c6, 942c7), 
for in Plato's state any citizen may be elected to administer the laws 
which are the only permanent rulers.3 Far from following Sparta in this, 
Plato, as we have seen, castigated the Spartan and Cretan practice of 
directing all the energies of the state towards preparation for war.' The 
greatest good is neither war nor civil strife ... but peace and goodwill.' 
'No one who makes warfare his sole or first concern can be a true 
statesman, nor legislate correctly unless his laws about war are designed 
to ensure peace, not his peacetime legislation as an instrument of war.' 
F28 c, d.) People say they wage war for the sake of peace, but its 
results are not what they hoped. Genuine leisure and culture (on which, 
he says, we chiefly set store) are never a consequence of war. One 
should spend one's life in play of the right sort, sacrificing, singing and 
dancing, and so win the favour o